Merge branch 'master' into for-next

Conflicts:
	arch/arm/mach-omap1/board-nokia770.c

146 files changed, 4203 insertions, 2938 deletions

diff --git a/fs/afs/Kconfig b/fs/afs/Kconfig
index 5c4e61d3c77..8f975f25b48 100644
--- a/fs/afs/Kconfig
+++ b/fs/afs/Kconfig
@@ -2,6 +2,7 @@ config AFS_FS
 	tristate "Andrew File System support (AFS) (EXPERIMENTAL)"
 	depends on INET && EXPERIMENTAL
 	select AF_RXRPC
+	select DNS_RESOLVER
 	help
 	  If you say Y here, you will get an experimental Andrew File System
 	  driver. It currently only supports unsecured read-only AFS access.
diff --git a/fs/afs/cell.c b/fs/afs/cell.c
index e19c13f059e..ffea35c6387 100644
--- a/fs/afs/cell.c
+++ b/fs/afs/cell.c
@@ -13,6 +13,7 @@
 #include <linux/slab.h>
 #include <linux/key.h>
 #include <linux/ctype.h>
+#include <linux/dns_resolver.h>
 #include <linux/sched.h>
 #include <keys/rxrpc-type.h>
 #include "internal.h"
@@ -36,6 +37,8 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist)
 	struct key *key;
 	size_t namelen;
 	char keyname[4 + AFS_MAXCELLNAME + 1], *cp, *dp, *next;
+	char  *dvllist = NULL, *_vllist = NULL;
+	char  delimiter = ':';
 	int ret;
 
 	_enter("%s,%s", name, vllist);
@@ -43,8 +46,10 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist)
 	BUG_ON(!name); /* TODO: want to look up "this cell" in the cache */
 
 	namelen = strlen(name);
-	if (namelen > AFS_MAXCELLNAME)
+	if (namelen > AFS_MAXCELLNAME) {
+		_leave(" = -ENAMETOOLONG");
 		return ERR_PTR(-ENAMETOOLONG);
+	}
 
 	/* allocate and initialise a cell record */
 	cell = kzalloc(sizeof(struct afs_cell) + namelen + 1, GFP_KERNEL);
@@ -64,15 +69,31 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist)
 	INIT_LIST_HEAD(&cell->vl_list);
 	spin_lock_init(&cell->vl_lock);
 
+	/* if the ip address is invalid, try dns query */
+	if (!vllist || strlen(vllist) < 7) {
+		ret = dns_query("afsdb", name, namelen, "ipv4", &dvllist, NULL);
+		if (ret < 0) {
+			_leave(" = %d", ret);
+			return ERR_PTR(ret);
+		}
+		_vllist = dvllist;
+
+		/* change the delimiter for user-space reply */
+		delimiter = ',';
+
+	} else {
+		_vllist = vllist;
+	}
+
 	/* fill in the VL server list from the rest of the string */
 	do {
 		unsigned a, b, c, d;
 
-		next = strchr(vllist, ':');
+		next = strchr(_vllist, delimiter);
 		if (next)
 			*next++ = 0;
 
-		if (sscanf(vllist, "%u.%u.%u.%u", &a, &b, &c, &d) != 4)
+		if (sscanf(_vllist, "%u.%u.%u.%u", &a, &b, &c, &d) != 4)
 			goto bad_address;
 
 		if (a > 255 || b > 255 || c > 255 || d > 255)
@@ -81,7 +102,7 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist)
 		cell->vl_addrs[cell->vl_naddrs++].s_addr =
 			htonl((a << 24) | (b << 16) | (c << 8) | d);
 
-	} while (cell->vl_naddrs < AFS_CELL_MAX_ADDRS && (vllist = next));
+	} while (cell->vl_naddrs < AFS_CELL_MAX_ADDRS && (_vllist = next));
 
 	/* create a key to represent an anonymous user */
 	memcpy(keyname, "afs@", 4);
@@ -110,6 +131,7 @@ bad_address:
 	ret = -EINVAL;
 error:
 	key_put(cell->anonymous_key);
+	kfree(dvllist);
 	kfree(cell);
 	_leave(" = %d", ret);
 	return ERR_PTR(ret);
@@ -201,14 +223,12 @@ int afs_cell_init(char *rootcell)
 	}
 
 	cp = strchr(rootcell, ':');
-	if (!cp) {
-		printk(KERN_ERR "kAFS: no VL server IP addresses specified\n");
-		_leave(" = -EINVAL");
-		return -EINVAL;
-	}
+	if (!cp)
+		_debug("kAFS: no VL server IP addresses specified");
+	else
+		*cp++ = 0;
 
 	/* allocate a cell record for the root cell */
-	*cp++ = 0;
 	new_root = afs_cell_create(rootcell, cp);
 	if (IS_ERR(new_root)) {
 		_leave(" = %ld", PTR_ERR(new_root));
diff --git a/fs/afs/main.c b/fs/afs/main.c
index 66d54d348c5..cfd1cbe25b2 100644
--- a/fs/afs/main.c
+++ b/fs/afs/main.c
@@ -111,6 +111,8 @@ static int __init afs_init(void)
 
 	/* initialise the callback update process */
 	ret = afs_callback_update_init();
+	if (ret < 0)
+		goto error_callback_update_init;
 
 	/* create the RxRPC transport */
 	ret = afs_open_socket();
@@ -127,15 +129,16 @@ static int __init afs_init(void)
 error_fs:
 	afs_close_socket();
 error_open_socket:
+	afs_callback_update_kill();
+error_callback_update_init:
+	afs_vlocation_purge();
 error_vl_update_init:
+	afs_cell_purge();
 error_cell_init:
 #ifdef CONFIG_AFS_FSCACHE
 	fscache_unregister_netfs(&afs_cache_netfs);
 error_cache:
 #endif
-	afs_callback_update_kill();
-	afs_vlocation_purge();
-	afs_cell_purge();
 	afs_proc_cleanup();
 	rcu_barrier();
 	printk(KERN_ERR "kAFS: failed to register: %d\n", ret);
diff --git a/fs/aio.c b/fs/aio.c
index 1ccf25cef1f..3006b5bc33d 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -1277,7 +1277,7 @@ out:
 /* sys_io_destroy:
  *	Destroy the aio_context specified.  May cancel any outstanding 
  *	AIOs and block on completion.  Will fail with -ENOSYS if not
- *	implemented.  May fail with -EFAULT if the context pointed to
+ *	implemented.  May fail with -EINVAL if the context pointed to
  *	is invalid.
  */
 SYSCALL_DEFINE1(io_destroy, aio_context_t, ctx)
@@ -1795,15 +1795,16 @@ SYSCALL_DEFINE3(io_cancel, aio_context_t, ctx_id, struct iocb __user *, iocb,
 
 /* io_getevents:
  *	Attempts to read at least min_nr events and up to nr events from
- *	the completion queue for the aio_context specified by ctx_id.  May
- *	fail with -EINVAL if ctx_id is invalid, if min_nr is out of range,
- *	if nr is out of range, if when is out of range.  May fail with
- *	-EFAULT if any of the memory specified to is invalid.  May return
- *	0 or < min_nr if no events are available and the timeout specified
- *	by when	has elapsed, where when == NULL specifies an infinite
- *	timeout.  Note that the timeout pointed to by when is relative and
- *	will be updated if not NULL and the operation blocks.  Will fail
- *	with -ENOSYS if not implemented.
+ *	the completion queue for the aio_context specified by ctx_id. If
+ *	it succeeds, the number of read events is returned. May fail with
+ *	-EINVAL if ctx_id is invalid, if min_nr is out of range, if nr is
+ *	out of range, if timeout is out of range.  May fail with -EFAULT
+ *	if any of the memory specified is invalid.  May return 0 or
+ *	< min_nr if the timeout specified by timeout has elapsed
+ *	before sufficient events are available, where timeout == NULL
+ *	specifies an infinite timeout. Note that the timeout pointed to by
+ *	timeout is relative and will be updated if not NULL and the
+ *	operation blocks. Will fail with -ENOSYS if not implemented.
  */
 SYSCALL_DEFINE5(io_getevents, aio_context_t, ctx_id,
 		long, min_nr,
diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c
index f4a7840bf42..42c7fafc8bf 100644
--- a/fs/cachefiles/namei.c
+++ b/fs/cachefiles/namei.c
@@ -37,9 +37,9 @@ void __cachefiles_printk_object(struct cachefiles_object *object,
 
 	printk(KERN_ERR "%sobject: OBJ%x\n",
 	       prefix, object->fscache.debug_id);
-	printk(KERN_ERR "%sobjstate=%s fl=%lx swfl=%lx ev=%lx[%lx]\n",
+	printk(KERN_ERR "%sobjstate=%s fl=%lx wbusy=%x ev=%lx[%lx]\n",
 	       prefix, fscache_object_states[object->fscache.state],
-	       object->fscache.flags, object->fscache.work.flags,
+	       object->fscache.flags, work_busy(&object->fscache.work),
 	       object->fscache.events,
 	       object->fscache.event_mask & FSCACHE_OBJECT_EVENTS_MASK);
 	printk(KERN_ERR "%sops=%u inp=%u exc=%u\n",
@@ -212,7 +212,7 @@ wait_for_old_object:
 
 		/* if the object we're waiting for is queued for processing,
 		 * then just put ourselves on the queue behind it */
-		if (slow_work_is_queued(&xobject->fscache.work)) {
+		if (work_pending(&xobject->fscache.work)) {
 			_debug("queue OBJ%x behind OBJ%x immediately",
 			       object->fscache.debug_id,
 			       xobject->fscache.debug_id);
@@ -220,8 +220,7 @@ wait_for_old_object:
 		}
 
 		/* otherwise we sleep until either the object we're waiting for
-		 * is done, or the slow-work facility wants the thread back to
-		 * do other work */
+		 * is done, or the fscache_object is congested */
 		wq = bit_waitqueue(&xobject->flags, CACHEFILES_OBJECT_ACTIVE);
 		init_wait(&wait);
 		requeue = false;
@@ -229,8 +228,8 @@ wait_for_old_object:
 			prepare_to_wait(wq, &wait, TASK_UNINTERRUPTIBLE);
 			if (!test_bit(CACHEFILES_OBJECT_ACTIVE, &xobject->flags))
 				break;
-			requeue = slow_work_sleep_till_thread_needed(
-				&object->fscache.work, &timeout);
+
+			requeue = fscache_object_sleep_till_congested(&timeout);
 		} while (timeout > 0 && !requeue);
 		finish_wait(wq, &wait);
 
diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c
index 0f0d41fbb03..0e3c0924cc3 100644
--- a/fs/cachefiles/rdwr.c
+++ b/fs/cachefiles/rdwr.c
@@ -422,7 +422,7 @@ int cachefiles_read_or_alloc_page(struct fscache_retrieval *op,
 	shift = PAGE_SHIFT - inode->i_sb->s_blocksize_bits;
 
 	op->op.flags &= FSCACHE_OP_KEEP_FLAGS;
-	op->op.flags |= FSCACHE_OP_FAST;
+	op->op.flags |= FSCACHE_OP_ASYNC;
 	op->op.processor = cachefiles_read_copier;
 
 	pagevec_init(&pagevec, 0);
@@ -729,7 +729,7 @@ int cachefiles_read_or_alloc_pages(struct fscache_retrieval *op,
 	pagevec_init(&pagevec, 0);
 
 	op->op.flags &= FSCACHE_OP_KEEP_FLAGS;
-	op->op.flags |= FSCACHE_OP_FAST;
+	op->op.flags |= FSCACHE_OP_ASYNC;
 	op->op.processor = cachefiles_read_copier;
 
 	INIT_LIST_HEAD(&backpages);
diff --git a/fs/char_dev.c b/fs/char_dev.c
index d6db933df2b..f80a4f25123 100644
--- a/fs/char_dev.c
+++ b/fs/char_dev.c
@@ -20,6 +20,7 @@
 #include <linux/cdev.h>
 #include <linux/mutex.h>
 #include <linux/backing-dev.h>
+#include <linux/tty.h>
 
 #include "internal.h"
 
diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig
index 5739fd7f88b..917b7d449bb 100644
--- a/fs/cifs/Kconfig
+++ b/fs/cifs/Kconfig
@@ -2,7 +2,6 @@ config CIFS
 	tristate "CIFS support (advanced network filesystem, SMBFS successor)"
 	depends on INET
 	select NLS
-	select SLOW_WORK
 	help
 	  This is the client VFS module for the Common Internet File System
 	  (CIFS) protocol which is the successor to the Server Message Block
@@ -71,14 +70,14 @@ config CIFS_WEAK_PW_HASH
 	  If unsure, say N.
 
 config CIFS_UPCALL
-	  bool "Kerberos/SPNEGO advanced session setup"
-	  depends on CIFS && KEYS
-	  help
-	    Enables an upcall mechanism for CIFS which accesses
-	    userspace helper utilities to provide SPNEGO packaged (RFC 4178)
-	    Kerberos tickets which are needed to mount to certain secure servers
-	    (for which more secure Kerberos authentication is required). If
-	    unsure, say N.
+	bool "Kerberos/SPNEGO advanced session setup"
+	depends on CIFS && KEYS
+	select DNS_RESOLVER
+	help
+	  Enables an upcall mechanism for CIFS which accesses userspace helper
+	  utilities to provide SPNEGO packaged (RFC 4178) Kerberos tickets
+	  which are needed to mount to certain secure servers (for which more
+	  secure Kerberos authentication is required). If unsure, say N.
 
 config CIFS_XATTR
         bool "CIFS extended attributes"
@@ -122,6 +121,7 @@ config CIFS_DEBUG2
 config CIFS_DFS_UPCALL
 	  bool "DFS feature support"
 	  depends on CIFS && KEYS
+	  select DNS_RESOLVER
 	  help
 	    Distributed File System (DFS) support is used to access shares
 	    transparently in an enterprise name space, even if the share
diff --git a/fs/cifs/README b/fs/cifs/README
index a727b7cb075..a7081eeeb85 100644
--- a/fs/cifs/README
+++ b/fs/cifs/README
@@ -568,8 +568,9 @@ module can be displayed via modinfo.
 Misc /proc/fs/cifs Flags and Debug Info
 =======================================
 Informational pseudo-files:
-DebugData		Displays information about active CIFS sessions
-			and shares, as well as the cifs.ko version.
+DebugData		Displays information about active CIFS sessions and
+			shares, features enabled as well as the cifs.ko
+			version.
 Stats			Lists summary resource usage information as well as per
 			share statistics, if CONFIG_CIFS_STATS in enabled
 			in the kernel configuration.
diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c
index 4fce6e61b34..eb1ba493489 100644
--- a/fs/cifs/cifs_debug.c
+++ b/fs/cifs/cifs_debug.c
@@ -119,6 +119,31 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v)
 		    "Display Internal CIFS Data Structures for Debugging\n"
 		    "---------------------------------------------------\n");
 	seq_printf(m, "CIFS Version %s\n", CIFS_VERSION);
+	seq_printf(m, "Features: ");
+#ifdef CONFIG_CIFS_DFS_UPCALL
+	seq_printf(m, "dfs");
+	seq_putc(m, ' ');
+#endif
+#ifdef CONFIG_CIFS_FSCACHE
+	seq_printf(m, "fscache");
+	seq_putc(m, ' ');
+#endif
+#ifdef CONFIG_CIFS_WEAK_PW_HASH
+	seq_printf(m, "lanman");
+	seq_putc(m, ' ');
+#endif
+#ifdef CONFIG_CIFS_POSIX
+	seq_printf(m, "posix");
+	seq_putc(m, ' ');
+#endif
+#ifdef CONFIG_CIFS_UPCALL
+	seq_printf(m, "spnego");
+	seq_putc(m, ' ');
+#endif
+#ifdef CONFIG_CIFS_XATTR
+	seq_printf(m, "xattr");
+#endif
+	seq_putc(m, '\n');
 	seq_printf(m, "Active VFS Requests: %d\n", GlobalTotalActiveXid);
 	seq_printf(m, "Servers:");
 
diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c
index dc1ed50ea06..d6ced7aa23c 100644
--- a/fs/cifs/cifs_dfs_ref.c
+++ b/fs/cifs/cifs_dfs_ref.c
@@ -141,7 +141,7 @@ char *cifs_compose_mount_options(const char *sb_mountdata,
 	}
 
 	rc = dns_resolve_server_name_to_ip(*devname, &srvIP);
-	if (rc != 0) {
+	if (rc < 0) {
 		cERROR(1, "%s: Failed to resolve server part of %s to IP: %d",
 			  __func__, *devname, rc);
 		goto compose_mount_options_err;
@@ -150,8 +150,7 @@ char *cifs_compose_mount_options(const char *sb_mountdata,
 	 * assuming that we have 'unc=' and 'ip=' in
 	 * the original sb_mountdata
 	 */
-	md_len = strlen(sb_mountdata) + strlen(srvIP) +
-		strlen(ref->node_name) + 12;
+	md_len = strlen(sb_mountdata) + rc + strlen(ref->node_name) + 12;
 	mountdata = kzalloc(md_len+1, GFP_KERNEL);
 	if (mountdata == NULL) {
 		rc = -ENOMEM;
diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c
index 6effccff85a..87044906cd1 100644
--- a/fs/cifs/cifs_spnego.c
+++ b/fs/cifs/cifs_spnego.c
@@ -84,6 +84,9 @@ struct key_type cifs_spnego_key_type = {
 /* strlen of ";uid=0x" */
 #define UID_KEY_LEN		7
 
+/* strlen of ";creduid=0x" */
+#define CREDUID_KEY_LEN		11
+
 /* strlen of ";user=" */
 #define USER_KEY_LEN		6
 
@@ -107,6 +110,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo)
 		   IP_KEY_LEN + INET6_ADDRSTRLEN +
 		   MAX_MECH_STR_LEN +
 		   UID_KEY_LEN + (sizeof(uid_t) * 2) +
+		   CREDUID_KEY_LEN + (sizeof(uid_t) * 2) +
 		   USER_KEY_LEN + strlen(sesInfo->userName) +
 		   PID_KEY_LEN + (sizeof(pid_t) * 2) + 1;
 
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index 8a2cf129e53..a5ed10c9afe 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -45,7 +45,6 @@
 #include "cifs_fs_sb.h"
 #include <linux/mm.h>
 #include <linux/key-type.h>
-#include "dns_resolve.h"
 #include "cifs_spnego.h"
 #include "fscache.h"
 #define CIFS_MAGIC_NUMBER 0xFF534D42	/* the first four bytes of SMB PDUs */
@@ -934,27 +933,13 @@ init_cifs(void)
 	if (rc)
 		goto out_unregister_filesystem;
 #endif
-#ifdef CONFIG_CIFS_DFS_UPCALL
-	rc = cifs_init_dns_resolver();
-	if (rc)
-		goto out_unregister_key_type;
-#endif
-	rc = slow_work_register_user(THIS_MODULE);
-	if (rc)
-		goto out_unregister_resolver_key;
 
 	return 0;
 
- out_unregister_resolver_key:
-#ifdef CONFIG_CIFS_DFS_UPCALL
-	cifs_exit_dns_resolver();
- out_unregister_key_type:
-#endif
 #ifdef CONFIG_CIFS_UPCALL
-	unregister_key_type(&cifs_spnego_key_type);
  out_unregister_filesystem:
-#endif
 	unregister_filesystem(&cifs_fs_type);
+#endif
  out_destroy_request_bufs:
 	cifs_destroy_request_bufs();
  out_destroy_mids:
@@ -976,7 +961,6 @@ exit_cifs(void)
 	cifs_fscache_unregister();
 #ifdef CONFIG_CIFS_DFS_UPCALL
 	cifs_dfs_release_automount_timer();
-	cifs_exit_dns_resolver();
 #endif
 #ifdef CONFIG_CIFS_UPCALL
 	unregister_key_type(&cifs_spnego_key_type);
diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h
index 59906146ad3..0cdfb8c32ac 100644
--- a/fs/cifs/cifsglob.h
+++ b/fs/cifs/cifsglob.h
@@ -22,7 +22,7 @@
 #include <linux/in.h>
 #include <linux/in6.h>
 #include <linux/slab.h>
-#include <linux/slow-work.h>
+#include <linux/workqueue.h>
 #include "cifs_fs_sb.h"
 #include "cifsacl.h"
 /*
@@ -356,7 +356,7 @@ struct cifsFileInfo {
 	atomic_t count;		/* reference count */
 	struct mutex fh_mutex; /* prevents reopen race after dead ses*/
 	struct cifs_search_info srch_inf;
-	struct slow_work oplock_break; /* slow_work job for oplock breaks */
+	struct work_struct oplock_break; /* work for oplock breaks */
 };
 
 /* Take a reference on the file private data */
@@ -728,6 +728,10 @@ GLOBAL_EXTERN unsigned int cifs_min_rcv;    /* min size of big ntwrk buf pool */
 GLOBAL_EXTERN unsigned int cifs_min_small;  /* min size of small buf pool */
 GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/
 
+void cifs_oplock_break(struct work_struct *work);
+void cifs_oplock_break_get(struct cifsFileInfo *cfile);
+void cifs_oplock_break_put(struct cifsFileInfo *cfile);
+
 extern const struct slow_work_ops cifs_oplock_break_ops;
 
 #endif	/* _CIFS_GLOB_H */
diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h
index 2eaebbd3113..1f545081408 100644
--- a/fs/cifs/cifsproto.h
+++ b/fs/cifs/cifsproto.h
@@ -86,8 +86,8 @@ extern unsigned int smbCalcSize(struct smb_hdr *ptr);
 extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr);
 extern int decode_negTokenInit(unsigned char *security_blob, int length,
 			struct TCP_Server_Info *server);
-extern int cifs_convert_address(struct sockaddr *dst, char *src);
-extern int cifs_fill_sockaddr(struct sockaddr *dst, char *src,
+extern int cifs_convert_address(struct sockaddr *dst, const char *src, int len);
+extern int cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len,
 				unsigned short int port);
 extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr);
 extern void header_assemble(struct smb_hdr *, char /* command */ ,
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 2a43a0aca96..95c2ea67edf 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -1543,6 +1543,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info)
 	if (volume_info->UNCip && volume_info->UNC) {
 		rc = cifs_fill_sockaddr((struct sockaddr *)&addr,
 					volume_info->UNCip,
+					strlen(volume_info->UNCip),
 					volume_info->port);
 		if (!rc) {
 			/* we failed translating address */
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index a7de5e9fff1..578d88c5b46 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -157,7 +157,7 @@ cifs_new_fileinfo(struct inode *newinode, __u16 fileHandle,
 	mutex_init(&pCifsFile->lock_mutex);
 	INIT_LIST_HEAD(&pCifsFile->llist);
 	atomic_set(&pCifsFile->count, 1);
-	slow_work_init(&pCifsFile->oplock_break, &cifs_oplock_break_ops);
+	INIT_WORK(&pCifsFile->oplock_break, cifs_oplock_break);
 
 	write_lock(&GlobalSMBSeslock);
 	list_add(&pCifsFile->tlist, &cifs_sb->tcon->openFileList);
diff --git a/fs/cifs/dns_resolve.c b/fs/cifs/dns_resolve.c
index 3ad7f4300c4..0eb87026cad 100644
--- a/fs/cifs/dns_resolve.c
+++ b/fs/cifs/dns_resolve.c
@@ -4,6 +4,8 @@
  *   Copyright (c) 2007 Igor Mammedov
  *   Author(s): Igor Mammedov (niallain@gmail.com)
  *              Steve French (sfrench@us.ibm.com)
+ *              Wang Lei (wang840925@gmail.com)
+ *		David Howells (dhowells@redhat.com)
  *
  *   Contains the CIFS DFS upcall routines used for hostname to
  *   IP address translation.
@@ -24,214 +26,73 @@
  */
 
 #include <linux/slab.h>
-#include <linux/keyctl.h>
-#include <linux/key-type.h>
-#include <keys/user-type.h>
+#include <linux/dns_resolver.h>
 #include "dns_resolve.h"
 #include "cifsglob.h"
 #include "cifsproto.h"
 #include "cifs_debug.h"
 
-static const struct cred *dns_resolver_cache;
-
-/* Checks if supplied name is IP address
- * returns:
- * 		1 - name is IP
- * 		0 - name is not IP
- */
-static int
-is_ip(char *name)
-{
-	struct sockaddr_storage ss;
-
-	return cifs_convert_address((struct sockaddr *)&ss, name);
-}
-
-static int
-dns_resolver_instantiate(struct key *key, const void *data,
-		size_t datalen)
-{
-	int rc = 0;
-	char *ip;
-
-	ip = kmalloc(datalen + 1, GFP_KERNEL);
-	if (!ip)
-		return -ENOMEM;
-
-	memcpy(ip, data, datalen);
-	ip[datalen] = '\0';
-
-	/* make sure this looks like an address */
-	if (!is_ip(ip)) {
-		kfree(ip);
-		return -EINVAL;
-	}
-
-	key->type_data.x[0] = datalen;
-	key->payload.data = ip;
-
-	return rc;
-}
-
-static void
-dns_resolver_destroy(struct key *key)
-{
-	kfree(key->payload.data);
-}
-
-struct key_type key_type_dns_resolver = {
-	.name        = "dns_resolver",
-	.def_datalen = sizeof(struct in_addr),
-	.describe    = user_describe,
-	.instantiate = dns_resolver_instantiate,
-	.destroy     = dns_resolver_destroy,
-	.match       = user_match,
-};
-
-/* Resolves server name to ip address.
- * input:
- * 	unc - server UNC
- * output:
- * 	*ip_addr - pointer to server ip, caller responcible for freeing it.
- * return 0 on success
+/**
+ * dns_resolve_server_name_to_ip - Resolve UNC server name to ip address.
+ * @unc: UNC path specifying the server
+ * @ip_addr: Where to return the IP address.
+ *
+ * The IP address will be returned in string form, and the caller is
+ * responsible for freeing it.
+ *
+ * Returns length of result on success, -ve on error.
  */
 int
 dns_resolve_server_name_to_ip(const char *unc, char **ip_addr)
 {
-	const struct cred *saved_cred;
-	int rc = -EAGAIN;
-	struct key *rkey = ERR_PTR(-EAGAIN);
+	struct sockaddr_storage ss;
+	const char *hostname, *sep;
 	char *name;
-	char *data = NULL;
-	int len;
+	int len, rc;
 
 	if (!ip_addr || !unc)
 		return -EINVAL;
 
-	/* search for server name delimiter */
 	len = strlen(unc);
 	if (len < 3) {
 		cFYI(1, "%s: unc is too short: %s", __func__, unc);
 		return -EINVAL;
 	}
-	len -= 2;
-	name = memchr(unc+2, '\\', len);
-	if (!name) {
-		cFYI(1, "%s: probably server name is whole unc: %s",
-					__func__, unc);
-	} else {
-		len = (name - unc) - 2/* leading // */;
-	}
-
-	name = kmalloc(len+1, GFP_KERNEL);
-	if (!name) {
-		rc = -ENOMEM;
-		return rc;
-	}
-	memcpy(name, unc+2, len);
-	name[len] = 0;
-
-	if (is_ip(name)) {
-		cFYI(1, "%s: it is IP, skipping dns upcall: %s",
-					__func__, name);
-		data = name;
-		goto skip_upcall;
-	}
 
-	saved_cred = override_creds(dns_resolver_cache);
-	rkey = request_key(&key_type_dns_resolver, name, "");
-	revert_creds(saved_cred);
-	if (!IS_ERR(rkey)) {
-		if (!(rkey->perm & KEY_USR_VIEW)) {
-			down_read(&rkey->sem);
-			rkey->perm |= KEY_USR_VIEW;
-			up_read(&rkey->sem);
-		}
-		len = rkey->type_data.x[0];
-		data = rkey->payload.data;
-	} else {
-		cERROR(1, "%s: unable to resolve: %s", __func__, name);
-		goto out;
-	}
-
-skip_upcall:
-	if (data) {
-		*ip_addr = kmalloc(len + 1, GFP_KERNEL);
-		if (*ip_addr) {
-			memcpy(*ip_addr, data, len + 1);
-			if (!IS_ERR(rkey))
-				cFYI(1, "%s: resolved: %s to %s", __func__,
-							name,
-							*ip_addr
-					);
-			rc = 0;
-		} else {
-			rc = -ENOMEM;
-		}
-		if (!IS_ERR(rkey))
-			key_put(rkey);
-	}
+	/* Discount leading slashes for cifs */
+	len -= 2;
+	hostname = unc + 2;
 
-out:
-	kfree(name);
+	/* Search for server name delimiter */
+	sep = memchr(hostname, '\\', len);
+	if (sep)
+		len = sep - unc;
+	else
+		cFYI(1, "%s: probably server name is whole unc: %s",
+		     __func__, unc);
+
+	/* Try to interpret hostname as an IPv4 or IPv6 address */
+	rc = cifs_convert_address((struct sockaddr *)&ss, hostname, len);
+	if (rc > 0)
+		goto name_is_IP_address;
+
+	/* Perform the upcall */
+	rc = dns_query(NULL, hostname, len, NULL, ip_addr, NULL);
+	if (rc < 0)
+		cERROR(1, "%s: unable to resolve: %*.*s",
+		       __func__, len, len, hostname);
+	else
+		cFYI(1, "%s: resolved: %*.*s to %s",
+		     __func__, len, len, hostname, *ip_addr);
 	return rc;
-}
 
-int __init cifs_init_dns_resolver(void)
-{
-	struct cred *cred;
-	struct key *keyring;
-	int ret;
-
-	printk(KERN_NOTICE "Registering the %s key type\n",
-	       key_type_dns_resolver.name);
-
-	/* create an override credential set with a special thread keyring in
-	 * which DNS requests are cached
-	 *
-	 * this is used to prevent malicious redirections from being installed
-	 * with add_key().
-	 */
-	cred = prepare_kernel_cred(NULL);
-	if (!cred)
+name_is_IP_address:
+	name = kmalloc(len + 1, GFP_KERNEL);
+	if (!name)
 		return -ENOMEM;
-
-	keyring = key_alloc(&key_type_keyring, ".dns_resolver", 0, 0, cred,
-			    (KEY_POS_ALL & ~KEY_POS_SETATTR) |
-			    KEY_USR_VIEW | KEY_USR_READ,
-			    KEY_ALLOC_NOT_IN_QUOTA);
-	if (IS_ERR(keyring)) {
-		ret = PTR_ERR(keyring);
-		goto failed_put_cred;
-	}
-
-	ret = key_instantiate_and_link(keyring, NULL, 0, NULL, NULL);
-	if (ret < 0)
-		goto failed_put_key;
-
-	ret = register_key_type(&key_type_dns_resolver);
-	if (ret < 0)
-		goto failed_put_key;
-
-	/* instruct request_key() to use this special keyring as a cache for
-	 * the results it looks up */
-	cred->thread_keyring = keyring;
-	cred->jit_keyring = KEY_REQKEY_DEFL_THREAD_KEYRING;
-	dns_resolver_cache = cred;
+	memcpy(name, hostname, len);
+	name[len] = 0;
+	cFYI(1, "%s: unc is IP, skipping dns upcall: %s", __func__, name);
+	*ip_addr = name;
 	return 0;
-
-failed_put_key:
-	key_put(keyring);
-failed_put_cred:
-	put_cred(cred);
-	return ret;
-}
-
-void cifs_exit_dns_resolver(void)
-{
-	key_revoke(dns_resolver_cache->thread_keyring);
-	unregister_key_type(&key_type_dns_resolver);
-	put_cred(dns_resolver_cache);
-	printk(KERN_NOTICE "Unregistered %s key type\n",
-	       key_type_dns_resolver.name);
 }
diff --git a/fs/cifs/dns_resolve.h b/fs/cifs/dns_resolve.h
index 5d7f291df16..d3f5d27f4d0 100644
--- a/fs/cifs/dns_resolve.h
+++ b/fs/cifs/dns_resolve.h
@@ -24,8 +24,6 @@
 #define _DNS_RESOLVE_H
 
 #ifdef __KERNEL__
-extern int __init cifs_init_dns_resolver(void);
-extern void cifs_exit_dns_resolver(void);
 extern int dns_resolve_server_name_to_ip(const char *unc, char **ip_addr);
 #endif /* KERNEL */
 
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index fa04a00d126..db11fdef0e9 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -2307,8 +2307,7 @@ static void cifs_invalidate_page(struct page *page, unsigned long offset)
 		cifs_fscache_invalidate_page(page, &cifsi->vfs_inode);
 }
 
-static void
-cifs_oplock_break(struct slow_work *work)
+void cifs_oplock_break(struct work_struct *work)
 {
 	struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo,
 						  oplock_break);
@@ -2345,33 +2344,30 @@ cifs_oplock_break(struct slow_work *work)
 				 LOCKING_ANDX_OPLOCK_RELEASE, false);
 		cFYI(1, "Oplock release rc = %d", rc);
 	}
+
+	/*
+	 * We might have kicked in before is_valid_oplock_break()
+	 * finished grabbing reference for us.  Make sure it's done by
+	 * waiting for GlobalSMSSeslock.
+	 */
+	write_lock(&GlobalSMBSeslock);
+	write_unlock(&GlobalSMBSeslock);
+
+	cifs_oplock_break_put(cfile);
 }
 
-static int
-cifs_oplock_break_get(struct slow_work *work)
+void cifs_oplock_break_get(struct cifsFileInfo *cfile)
 {
-	struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo,
-						  oplock_break);
 	mntget(cfile->mnt);
 	cifsFileInfo_get(cfile);
-	return 0;
 }
 
-static void
-cifs_oplock_break_put(struct slow_work *work)
+void cifs_oplock_break_put(struct cifsFileInfo *cfile)
 {
-	struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo,
-						  oplock_break);
 	mntput(cfile->mnt);
 	cifsFileInfo_put(cfile);
 }
 
-const struct slow_work_ops cifs_oplock_break_ops = {
-	.get_ref	= cifs_oplock_break_get,
-	.put_ref	= cifs_oplock_break_put,
-	.execute	= cifs_oplock_break,
-};
-
 const struct address_space_operations cifs_addr_ops = {
 	.readpage = cifs_readpage,
 	.readpages = cifs_readpages,
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c
index a15b3a9bbff..dc4c47ab958 100644
--- a/fs/cifs/inode.c
+++ b/fs/cifs/inode.c
@@ -732,15 +732,9 @@ cifs_find_inode(struct inode *inode, void *opaque)
 	if ((inode->i_mode & S_IFMT) != (fattr->cf_mode & S_IFMT))
 		return 0;
 
-	/*
-	 * uh oh -- it's a directory. We can't use it since hardlinked dirs are
-	 * verboten. Disable serverino and return it as if it were found, the
-	 * caller can discard it, generate a uniqueid and retry the find
-	 */
-	if (S_ISDIR(inode->i_mode) && !list_empty(&inode->i_dentry)) {
+	/* if it's not a directory or has no dentries, then flag it */
+	if (S_ISDIR(inode->i_mode) && !list_empty(&inode->i_dentry))
 		fattr->cf_flags |= CIFS_FATTR_INO_COLLISION;
-		cifs_autodisable_serverino(CIFS_SB(inode->i_sb));
-	}
 
 	return 1;
 }
@@ -754,6 +748,27 @@ cifs_init_inode(struct inode *inode, void *opaque)
 	return 0;
 }
 
+/*
+ * walk dentry list for an inode and report whether it has aliases that
+ * are hashed. We use this to determine if a directory inode can actually
+ * be used.
+ */
+static bool
+inode_has_hashed_dentries(struct inode *inode)
+{
+	struct dentry *dentry;
+
+	spin_lock(&dcache_lock);
+	list_for_each_entry(dentry, &inode->i_dentry, d_alias) {
+		if (!d_unhashed(dentry) || IS_ROOT(dentry)) {
+			spin_unlock(&dcache_lock);
+			return true;
+		}
+	}
+	spin_unlock(&dcache_lock);
+	return false;
+}
+
 /* Given fattrs, get a corresponding inode */
 struct inode *
 cifs_iget(struct super_block *sb, struct cifs_fattr *fattr)
@@ -769,12 +784,16 @@ retry_iget5_locked:
 
 	inode = iget5_locked(sb, hash, cifs_find_inode, cifs_init_inode, fattr);
 	if (inode) {
-		/* was there a problematic inode number collision? */
+		/* was there a potentially problematic inode collision? */
 		if (fattr->cf_flags & CIFS_FATTR_INO_COLLISION) {
-			iput(inode);
-			fattr->cf_uniqueid = iunique(sb, ROOT_I);
 			fattr->cf_flags &= ~CIFS_FATTR_INO_COLLISION;
-			goto retry_iget5_locked;
+
+			if (inode_has_hashed_dentries(inode)) {
+				cifs_autodisable_serverino(CIFS_SB(sb));
+				iput(inode);
+				fattr->cf_uniqueid = iunique(sb, ROOT_I);
+				goto retry_iget5_locked;
+			}
 		}
 
 		cifs_fattr_to_inode(inode, fattr);
diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c
index 1394aa37f26..3ccadc1326d 100644
--- a/fs/cifs/misc.c
+++ b/fs/cifs/misc.c
@@ -498,7 +498,6 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv)
 	struct cifsTconInfo *tcon;
 	struct cifsInodeInfo *pCifsInode;
 	struct cifsFileInfo *netfile;
-	int rc;
 
 	cFYI(1, "Checking for oplock break or dnotify response");
 	if ((pSMB->hdr.Command == SMB_COM_NT_TRANSACT) &&
@@ -583,13 +582,18 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv)
 				pCifsInode->clientCanCacheAll = false;
 				if (pSMB->OplockLevel == 0)
 					pCifsInode->clientCanCacheRead = false;
-				rc = slow_work_enqueue(&netfile->oplock_break);
-				if (rc) {
-					cERROR(1, "failed to enqueue oplock "
-						   "break: %d\n", rc);
-				} else {
-					netfile->oplock_break_cancelled = false;
-				}
+
+				/*
+				 * cifs_oplock_break_put() can't be called
+				 * from here.  Get reference after queueing
+				 * succeeded.  cifs_oplock_break() will
+				 * synchronize using GlobalSMSSeslock.
+				 */
+				if (queue_work(system_nrt_wq,
+					       &netfile->oplock_break))
+					cifs_oplock_break_get(netfile);
+				netfile->oplock_break_cancelled = false;
+
 				read_unlock(&GlobalSMBSeslock);
 				read_unlock(&cifs_tcp_ses_lock);
 				return true;
diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c
index c6721ee26db..f97851119e6 100644
--- a/fs/cifs/netmisc.c
+++ b/fs/cifs/netmisc.c
@@ -140,17 +140,18 @@ static const struct smb_to_posix_error mapping_table_ERRHRD[] = {
  * Returns 0 on failure.
  */
 static int
-cifs_inet_pton(const int address_family, const char *cp, void *dst)
+cifs_inet_pton(const int address_family, const char *cp, int len, void *dst)
 {
 	int ret = 0;
 
 	/* calculate length by finding first slash or NULL */
 	if (address_family == AF_INET)
-		ret = in4_pton(cp, -1 /* len */, dst, '\\', NULL);
+		ret = in4_pton(cp, len, dst, '\\', NULL);
 	else if (address_family == AF_INET6)
-		ret = in6_pton(cp, -1 /* len */, dst , '\\', NULL);
+		ret = in6_pton(cp, len, dst , '\\', NULL);
 
-	cFYI(DBG2, "address conversion returned %d for %s", ret, cp);
+	cFYI(DBG2, "address conversion returned %d for %*.*s",
+	     ret, len, len, cp);
 	if (ret > 0)
 		ret = 1;
 	return ret;
@@ -165,37 +166,39 @@ cifs_inet_pton(const int address_family, const char *cp, void *dst)
  * Returns 0 on failure.
  */
 int
-cifs_convert_address(struct sockaddr *dst, char *src)
+cifs_convert_address(struct sockaddr *dst, const char *src, int len)
 {
-	int rc;
-	char *pct, *endp;
+	int rc, alen, slen;
+	const char *pct;
+	char *endp, scope_id[13];
 	struct sockaddr_in *s4 = (struct sockaddr_in *) dst;
 	struct sockaddr_in6 *s6 = (struct sockaddr_in6 *) dst;
 
 	/* IPv4 address */
-	if (cifs_inet_pton(AF_INET, src, &s4->sin_addr.s_addr)) {
+	if (cifs_inet_pton(AF_INET, src, len, &s4->sin_addr.s_addr)) {
 		s4->sin_family = AF_INET;
 		return 1;
 	}
 
-	/* temporarily terminate string */
-	pct = strchr(src, '%');
-	if (pct)
-		*pct = '\0';
-
-	rc = cifs_inet_pton(AF_INET6, src, &s6->sin6_addr.s6_addr);
-
-	/* repair temp termination (if any) and make pct point to scopeid */
-	if (pct)
-		*pct++ = '%';
+	/* attempt to exclude the scope ID from the address part */
+	pct = memchr(src, '%', len);
+	alen = pct ? pct - src : len;
 
+	rc = cifs_inet_pton(AF_INET6, src, alen, &s6->sin6_addr.s6_addr);
 	if (!rc)
 		return rc;
 
 	s6->sin6_family = AF_INET6;
 	if (pct) {
+		/* grab the scope ID */
+		slen = len - (alen + 1);
+		if (slen <= 0 || slen > 12)
+			return 0;
+		memcpy(scope_id, pct + 1, slen);
+		scope_id[slen] = '\0';
+
 		s6->sin6_scope_id = (u32) simple_strtoul(pct, &endp, 0);
-		if (!*pct || *endp)
+		if (endp != scope_id + slen)
 			return 0;
 	}
 
@@ -203,10 +206,10 @@ cifs_convert_address(struct sockaddr *dst, char *src)
 }
 
 int
-cifs_fill_sockaddr(struct sockaddr *dst, char *src,
+cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len,
 		   const unsigned short int port)
 {
-	if (!cifs_convert_address(dst, src))
+	if (!cifs_convert_address(dst, src, len))
 		return 0;
 
 	switch (dst->sa_family) {
diff --git a/fs/compat.c b/fs/compat.c
index c6fda9aeb86..5976bad85f6 100644
--- a/fs/compat.c
+++ b/fs/compat.c
@@ -15,6 +15,7 @@
  *  published by the Free Software Foundation.
  */
 
+#include <linux/stddef.h>
 #include <linux/kernel.h>
 #include <linux/linkage.h>
 #include <linux/compat.h>
@@ -891,8 +892,6 @@ asmlinkage long compat_sys_mount(char __user * dev_name, char __user * dir_name,
 	return retval;
 }
 
-#define NAME_OFFSET(de) ((int) ((de)->d_name - (char __user *) (de)))
-
 struct compat_old_linux_dirent {
 	compat_ulong_t	d_ino;
 	compat_ulong_t	d_offset;
@@ -981,7 +980,8 @@ static int compat_filldir(void *__buf, const char *name, int namlen,
 	struct compat_linux_dirent __user * dirent;
 	struct compat_getdents_callback *buf = __buf;
 	compat_ulong_t d_ino;
-	int reclen = ALIGN(NAME_OFFSET(dirent) + namlen + 2, sizeof(compat_long_t));
+	int reclen = ALIGN(offsetof(struct compat_linux_dirent, d_name) +
+		namlen + 2, sizeof(compat_long_t));
 
 	buf->error = -EINVAL;	/* only used if we fail.. */
 	if (reclen > buf->count)
@@ -1068,8 +1068,8 @@ static int compat_filldir64(void * __buf, const char * name, int namlen, loff_t
 {
 	struct linux_dirent64 __user *dirent;
 	struct compat_getdents_callback64 *buf = __buf;
-	int jj = NAME_OFFSET(dirent);
-	int reclen = ALIGN(jj + namlen + 1, sizeof(u64));
+	int reclen = ALIGN(offsetof(struct linux_dirent64, d_name) + namlen + 1,
+		sizeof(u64));
 	u64 off;
 
 	buf->error = -EINVAL;	/* only used if we fail.. */
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index c0d35c62052..37a34c2c622 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -248,7 +248,7 @@ static struct connection *assoc2con(int assoc_id)
 
 	for (i = 0 ; i < CONN_HASH_SIZE; i++) {
 		hlist_for_each_entry(con, h, &connection_hash[i], list) {
-			if (con && con->sctp_assoc == assoc_id) {
+			if (con->sctp_assoc == assoc_id) {
 				mutex_unlock(&connections_lock);
 				return con;
 			}
diff --git a/fs/dlm/netlink.c b/fs/dlm/netlink.c
index 2c6ad518100..ef17e0169da 100644
--- a/fs/dlm/netlink.c
+++ b/fs/dlm/netlink.c
@@ -81,24 +81,11 @@ static struct genl_ops dlm_nl_ops = {
 
 int __init dlm_netlink_init(void)
 {
-	int rv;
-
-	rv = genl_register_family(&family);
-	if (rv)
-		return rv;
-
-	rv = genl_register_ops(&family, &dlm_nl_ops);
-	if (rv < 0)
-		goto err;
-	return 0;
- err:
-	genl_unregister_family(&family);
-	return rv;
+	return genl_register_family_with_ops(&family, &dlm_nl_ops, 1);
 }
 
 void dlm_netlink_exit(void)
 {
-	genl_unregister_ops(&family, &dlm_nl_ops);
 	genl_unregister_family(&family);
 }
 
diff --git a/fs/exec.c b/fs/exec.c
index e19de6a8033..dab85ecad68 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -28,7 +28,6 @@
 #include <linux/mm.h>
 #include <linux/stat.h>
 #include <linux/fcntl.h>
-#include <linux/smp_lock.h>
 #include <linux/swap.h>
 #include <linux/string.h>
 #include <linux/init.h>
@@ -653,6 +652,7 @@ int setup_arg_pages(struct linux_binprm *bprm,
 	else
 		stack_base = vma->vm_start - stack_expand;
 #endif
+	current->mm->start_stack = bprm->p;
 	ret = expand_stack(vma, stack_base);
 	if (ret)
 		ret = -EFAULT;
@@ -1891,13 +1891,7 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs)
 	 */
 	clear_thread_flag(TIF_SIGPENDING);
 
-	/*
-	 * lock_kernel() because format_corename() is controlled by sysctl, which
-	 * uses lock_kernel()
-	 */
- 	lock_kernel();
 	ispipe = format_corename(corename, signr);
-	unlock_kernel();
 
  	if (ispipe) {
 		int dump_count;
diff --git a/fs/ext3/Kconfig b/fs/ext3/Kconfig
index 522b15498f4..e8c6ba0e4a3 100644
--- a/fs/ext3/Kconfig
+++ b/fs/ext3/Kconfig
@@ -31,6 +31,7 @@ config EXT3_FS
 config EXT3_DEFAULTS_TO_ORDERED
 	bool "Default to 'data=ordered' in ext3"
 	depends on EXT3_FS
+	default y
 	help
 	  The journal mode options for ext3 have different tradeoffs
 	  between when data is guaranteed to be on disk and
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 735f0190ec2..001eb0e2d48 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -1149,9 +1149,25 @@ static int walk_page_buffers(	handle_t *handle,
 static int do_journal_get_write_access(handle_t *handle,
 					struct buffer_head *bh)
 {
+	int dirty = buffer_dirty(bh);
+	int ret;
+
 	if (!buffer_mapped(bh) || buffer_freed(bh))
 		return 0;
-	return ext3_journal_get_write_access(handle, bh);
+	/*
+	 * __block_prepare_write() could have dirtied some buffers. Clean
+	 * the dirty bit as jbd2_journal_get_write_access() could complain
+	 * otherwise about fs integrity issues. Setting of the dirty bit
+	 * by __block_prepare_write() isn't a real problem here as we clear
+	 * the bit before releasing a page lock and thus writeback cannot
+	 * ever write the buffer.
+	 */
+	if (dirty)
+		clear_buffer_dirty(bh);
+	ret = ext3_journal_get_write_access(handle, bh);
+	if (!ret && dirty)
+		ret = ext3_journal_dirty_metadata(handle, bh);
+	return ret;
 }
 
 /*
@@ -1625,10 +1641,7 @@ static int ext3_writeback_writepage(struct page *page,
 		goto out_fail;
 	}
 
-	if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode))
-		ret = nobh_writepage(page, ext3_get_block, wbc);
-	else
-		ret = block_write_full_page(page, ext3_get_block, wbc);
+	ret = block_write_full_page(page, ext3_get_block, wbc);
 
 	err = ext3_journal_stop(handle);
 	if (!ret)
@@ -1922,17 +1935,6 @@ static int ext3_block_truncate_page(handle_t *handle, struct page *page,
 	length = blocksize - (offset & (blocksize - 1));
 	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);
 
-	/*
-	 * For "nobh" option,  we can only work if we don't need to
-	 * read-in the page - otherwise we create buffers to do the IO.
-	 */
-	if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) &&
-	     ext3_should_writeback_data(inode) && PageUptodate(page)) {
-		zero_user(page, offset, length);
-		set_page_dirty(page);
-		goto unlock;
-	}
-
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, blocksize, 0);
 
@@ -2284,27 +2286,6 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode,
 					   depth);
 
 			/*
-			 * We've probably journalled the indirect block several
-			 * times during the truncate.  But it's no longer
-			 * needed and we now drop it from the transaction via
-			 * journal_revoke().
-			 *
-			 * That's easy if it's exclusively part of this
-			 * transaction.  But if it's part of the committing
-			 * transaction then journal_forget() will simply
-			 * brelse() it.  That means that if the underlying
-			 * block is reallocated in ext3_get_block(),
-			 * unmap_underlying_metadata() will find this block
-			 * and will try to get rid of it.  damn, damn.
-			 *
-			 * If this block has already been committed to the
-			 * journal, a revoke record will be written.  And
-			 * revoke records must be emitted *before* clearing
-			 * this block's bit in the bitmaps.
-			 */
-			ext3_forget(handle, 1, inode, bh, bh->b_blocknr);
-
-			/*
 			 * Everything below this this pointer has been
 			 * released.  Now let this top-of-subtree go.
 			 *
@@ -2327,6 +2308,31 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode,
 				truncate_restart_transaction(handle, inode);
 			}
 
+			/*
+			 * We've probably journalled the indirect block several
+			 * times during the truncate.  But it's no longer
+			 * needed and we now drop it from the transaction via
+			 * journal_revoke().
+			 *
+			 * That's easy if it's exclusively part of this
+			 * transaction.  But if it's part of the committing
+			 * transaction then journal_forget() will simply
+			 * brelse() it.  That means that if the underlying
+			 * block is reallocated in ext3_get_block(),
+			 * unmap_underlying_metadata() will find this block
+			 * and will try to get rid of it.  damn, damn. Thus
+			 * we don't allow a block to be reallocated until
+			 * a transaction freeing it has fully committed.
+			 *
+			 * We also have to make sure journal replay after a
+			 * crash does not overwrite non-journaled data blocks
+			 * with old metadata when the block got reallocated for
+			 * data.  Thus we have to store a revoke record for a
+			 * block in the same transaction in which we free the
+			 * block.
+			 */
+			ext3_forget(handle, 1, inode, bh, bh->b_blocknr);
+
 			ext3_free_blocks(handle, inode, nr, 1);
 
 			if (parent_bh) {
diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c
index ee184084ca4..2b35ddb70d6 100644
--- a/fs/ext3/namei.c
+++ b/fs/ext3/namei.c
@@ -1447,7 +1447,6 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry,
 	struct inode *inode)
 {
 	struct inode *dir = dentry->d_parent->d_inode;
-	unsigned long offset;
 	struct buffer_head * bh;
 	struct ext3_dir_entry_2 *de;
 	struct super_block * sb;
@@ -1469,7 +1468,7 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry,
 		ext3_mark_inode_dirty(handle, dir);
 	}
 	blocks = dir->i_size >> sb->s_blocksize_bits;
-	for (block = 0, offset = 0; block < blocks; block++) {
+	for (block = 0; block < blocks; block++) {
 		bh = ext3_bread(handle, dir, block, 0, &retval);
 		if(!bh)
 			return retval;
diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c
index 54351ac7cef..0ccd7b12b73 100644
--- a/fs/ext3/resize.c
+++ b/fs/ext3/resize.c
@@ -964,7 +964,6 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es,
 		      ext3_fsblk_t n_blocks_count)
 {
 	ext3_fsblk_t o_blocks_count;
-	unsigned long o_groups_count;
 	ext3_grpblk_t last;
 	ext3_grpblk_t add;
 	struct buffer_head * bh;
@@ -976,7 +975,6 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es,
 	 * yet: we're going to revalidate es->s_blocks_count after
 	 * taking the s_resize_lock below. */
 	o_blocks_count = le32_to_cpu(es->s_blocks_count);
-	o_groups_count = EXT3_SB(sb)->s_groups_count;
 
 	if (test_opt(sb, DEBUG))
 		printk(KERN_DEBUG "EXT3-fs: extending last group from "E3FSBLK" uto "E3FSBLK" blocks\n",
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index 6c953bb255e..9650a956fd0 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -661,9 +661,6 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	 */
 	seq_puts(seq, ",barrier=");
 	seq_puts(seq, test_opt(sb, BARRIER) ? "1" : "0");
-	if (test_opt(sb, NOBH))
-		seq_puts(seq, ",nobh");
-
 	seq_printf(seq, ",data=%s", data_mode_string(test_opt(sb, DATA_FLAGS)));
 	if (test_opt(sb, DATA_ERR_ABORT))
 		seq_puts(seq, ",data_err=abort");
@@ -1255,10 +1252,12 @@ set_qf_format:
 			*n_blocks_count = option;
 			break;
 		case Opt_nobh:
-			set_opt(sbi->s_mount_opt, NOBH);
+			ext3_msg(sb, KERN_WARNING,
+				"warning: ignoring deprecated nobh option");
 			break;
 		case Opt_bh:
-			clear_opt(sbi->s_mount_opt, NOBH);
+			ext3_msg(sb, KERN_WARNING,
+				"warning: ignoring deprecated bh option");
 			break;
 		default:
 			ext3_msg(sb, KERN_ERR,
@@ -2001,14 +2000,6 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent)
 		break;
 	}
 
-	if (test_opt(sb, NOBH)) {
-		if (!(test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA)) {
-			ext3_msg(sb, KERN_WARNING,
-				"warning: ignoring nobh option - "
-				"it is supported only with writeback mode");
-			clear_opt(sbi->s_mount_opt, NOBH);
-		}
-	}
 	/*
 	 * The journal_load will have done any necessary log recovery,
 	 * so we can safely mount the rest of the filesystem now.
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index feaf498feaa..5e2ed4504ea 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -204,6 +204,7 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type,
 				return error;
 			else {
 				inode->i_mode = mode;
+				inode->i_ctime = ext4_current_time(inode);
 				ext4_mark_inode_dirty(handle, inode);
 				if (error == 0)
 					acl = NULL;
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 95b7594c76f..bd30799a43e 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -377,14 +377,11 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb,
 	ext4_grpblk_t bit;
 	unsigned int i;
 	struct ext4_group_desc *desc;
-	struct ext4_super_block *es;
-	struct ext4_sb_info *sbi;
+	struct ext4_sb_info *sbi = EXT4_SB(sb);
 	int err = 0, ret, blk_free_count;
 	ext4_grpblk_t blocks_freed;
 	struct ext4_group_info *grp;
 
-	sbi = EXT4_SB(sb);
-	es = sbi->s_es;
 	ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1);
 
 	ext4_get_group_no_and_offset(sb, block, &block_group, &bit);
@@ -477,7 +474,6 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb,
 	ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh);
 	if (!err)
 		err = ret;
-	sb->s_dirt = 1;
 
 error_return:
 	brelse(bitmap_bh);
diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c
index 5b6973fbf1b..3db5084db9b 100644
--- a/fs/ext4/block_validity.c
+++ b/fs/ext4/block_validity.c
@@ -229,16 +229,20 @@ int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk,
 
 	if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) ||
 	    (start_blk + count < start_blk) ||
-	    (start_blk + count > ext4_blocks_count(sbi->s_es)))
+	    (start_blk + count > ext4_blocks_count(sbi->s_es))) {
+		sbi->s_es->s_last_error_block = cpu_to_le64(start_blk);
 		return 0;
+	}
 	while (n) {
 		entry = rb_entry(n, struct ext4_system_zone, node);
 		if (start_blk + count - 1 < entry->start_blk)
 			n = n->rb_left;
 		else if (start_blk >= (entry->start_blk + entry->count))
 			n = n->rb_right;
-		else
+		else {
+			sbi->s_es->s_last_error_block = cpu_to_le64(start_blk);
 			return 0;
+		}
 	}
 	return 1;
 }
diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index ea5e6cb7e2a..374510f72ba 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -61,10 +61,11 @@ static unsigned char get_dtype(struct super_block *sb, int filetype)
 }
 
 
-int ext4_check_dir_entry(const char *function, struct inode *dir,
-			 struct ext4_dir_entry_2 *de,
-			 struct buffer_head *bh,
-			 unsigned int offset)
+int __ext4_check_dir_entry(const char *function, unsigned int line,
+			   struct inode *dir,
+			   struct ext4_dir_entry_2 *de,
+			   struct buffer_head *bh,
+			   unsigned int offset)
 {
 	const char *error_msg = NULL;
 	const int rlen = ext4_rec_len_from_disk(de->rec_len,
@@ -83,11 +84,10 @@ int ext4_check_dir_entry(const char *function, struct inode *dir,
 		error_msg = "inode out of bounds";
 
 	if (error_msg != NULL)
-		ext4_error_inode(function, dir,
-			"bad entry in directory: %s - block=%llu"
+		ext4_error_inode(dir, function, line, bh->b_blocknr,
+			"bad entry in directory: %s - "
 			"offset=%u(%u), inode=%u, rec_len=%d, name_len=%d",
-			error_msg, (unsigned long long) bh->b_blocknr,
-			(unsigned) (offset%bh->b_size), offset,
+			error_msg, (unsigned) (offset%bh->b_size), offset,
 			le32_to_cpu(de->inode),
 			rlen, de->name_len);
 	return error_msg == NULL ? 1 : 0;
@@ -121,7 +121,8 @@ static int ext4_readdir(struct file *filp,
 		 * We don't set the inode dirty flag since it's not
 		 * critical that it get flushed back to the disk.
 		 */
-		ext4_clear_inode_flag(filp->f_path.dentry->d_inode, EXT4_INODE_INDEX);
+		ext4_clear_inode_flag(filp->f_path.dentry->d_inode,
+				      EXT4_INODE_INDEX);
 	}
 	stored = 0;
 	offset = filp->f_pos & (sb->s_blocksize - 1);
@@ -193,7 +194,7 @@ revalidate:
 		while (!error && filp->f_pos < inode->i_size
 		       && offset < sb->s_blocksize) {
 			de = (struct ext4_dir_entry_2 *) (bh->b_data + offset);
-			if (!ext4_check_dir_entry("ext4_readdir", inode, de,
+			if (!ext4_check_dir_entry(inode, de,
 						  bh, offset)) {
 				/*
 				 * On error, skip the f_pos to the next block
@@ -343,7 +344,7 @@ int ext4_htree_store_dirent(struct file *dir_file, __u32 hash,
 	struct dir_private_info *info;
 	int len;
 
-	info = (struct dir_private_info *) dir_file->private_data;
+	info = dir_file->private_data;
 	p = &info->root.rb_node;
 
 	/* Create and allocate the fname structure */
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 19a4de57128..e03841d9f30 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -57,10 +57,13 @@
 #endif
 
 #define EXT4_ERROR_INODE(inode, fmt, a...) \
-	ext4_error_inode(__func__, (inode), (fmt), ## a)
+	ext4_error_inode((inode), __func__, __LINE__, 0, (fmt), ## a)
+
+#define EXT4_ERROR_INODE_BLOCK(inode, block, fmt, a...)			\
+	ext4_error_inode((inode), __func__, __LINE__, (block), (fmt), ## a)
 
 #define EXT4_ERROR_FILE(file, fmt, a...)	\
-	ext4_error_file(__func__, (file), (fmt), ## a)
+	ext4_error_file(__func__, __LINE__, (file), (fmt), ## a)
 
 /* data type for block offset of block group */
 typedef int ext4_grpblk_t;
@@ -167,13 +170,15 @@ struct mpage_da_data {
 };
 #define	EXT4_IO_UNWRITTEN	0x1
 typedef struct ext4_io_end {
-	struct list_head	list;		/* per-file finished AIO list */
+	struct list_head	list;		/* per-file finished IO list */
 	struct inode		*inode;		/* file being written to */
 	unsigned int		flag;		/* unwritten or not */
 	struct page		*page;		/* page struct for buffer write */
 	loff_t			offset;		/* offset in the file */
 	ssize_t			size;		/* size of the extent */
 	struct work_struct	work;		/* data work queue */
+	struct kiocb		*iocb;		/* iocb struct for AIO */
+	int			result;		/* error value for AIO */
 } ext4_io_end_t;
 
 /*
@@ -460,7 +465,7 @@ struct ext4_new_group_data {
 };
 
 /*
- * Flags used by ext4_get_blocks()
+ * Flags used by ext4_map_blocks()
  */
 	/* Allocate any needed blocks and/or convert an unitialized
 	   extent to be an initialized ext4 */
@@ -873,7 +878,6 @@ struct ext4_inode_info {
 #define EXT4_MOUNT_POSIX_ACL		0x08000	/* POSIX Access Control Lists */
 #define EXT4_MOUNT_NO_AUTO_DA_ALLOC	0x10000	/* No auto delalloc mapping */
 #define EXT4_MOUNT_BARRIER		0x20000 /* Use block barriers */
-#define EXT4_MOUNT_NOBH			0x40000 /* No bufferheads */
 #define EXT4_MOUNT_QUOTA		0x80000 /* Some quota option set */
 #define EXT4_MOUNT_USRQUOTA		0x100000 /* "old" user quota */
 #define EXT4_MOUNT_GRPQUOTA		0x200000 /* "old" group quota */
@@ -982,7 +986,7 @@ struct ext4_super_block {
 	__le32	s_last_orphan;		/* start of list of inodes to delete */
 	__le32	s_hash_seed[4];		/* HTREE hash seed */
 	__u8	s_def_hash_version;	/* Default hash version to use */
-	__u8	s_reserved_char_pad;
+	__u8	s_jnl_backup_type;
 	__le16  s_desc_size;		/* size of group descriptor */
 /*100*/	__le32	s_default_mount_opts;
 	__le32	s_first_meta_bg;	/* First metablock block group */
@@ -1000,12 +1004,34 @@ struct ext4_super_block {
 	__le64  s_mmp_block;            /* Block for multi-mount protection */
 	__le32  s_raid_stripe_width;    /* blocks on all data disks (N*stride)*/
 	__u8	s_log_groups_per_flex;  /* FLEX_BG group size */
-	__u8	s_reserved_char_pad2;
+	__u8	s_reserved_char_pad;
 	__le16  s_reserved_pad;
 	__le64	s_kbytes_written;	/* nr of lifetime kilobytes written */
-	__u32   s_reserved[160];        /* Padding to the end of the block */
+	__le32	s_snapshot_inum;	/* Inode number of active snapshot */
+	__le32	s_snapshot_id;		/* sequential ID of active snapshot */
+	__le64	s_snapshot_r_blocks_count; /* reserved blocks for active
+					      snapshot's future use */
+	__le32	s_snapshot_list;	/* inode number of the head of the
+					   on-disk snapshot list */
+#define EXT4_S_ERR_START offsetof(struct ext4_super_block, s_error_count)
+	__le32	s_error_count;		/* number of fs errors */
+	__le32	s_first_error_time;	/* first time an error happened */
+	__le32	s_first_error_ino;	/* inode involved in first error */
+	__le64	s_first_error_block;	/* block involved of first error */
+	__u8	s_first_error_func[32];	/* function where the error happened */
+	__le32	s_first_error_line;	/* line number where error happened */
+	__le32	s_last_error_time;	/* most recent time of an error */
+	__le32	s_last_error_ino;	/* inode involved in last error */
+	__le32	s_last_error_line;	/* line number where error happened */
+	__le64	s_last_error_block;	/* block involved of last error */
+	__u8	s_last_error_func[32];	/* function where the error happened */
+#define EXT4_S_ERR_END offsetof(struct ext4_super_block, s_mount_opts)
+	__u8	s_mount_opts[64];
+	__le32	s_reserved[112];        /* Padding to the end of the block */
 };
 
+#define EXT4_S_ERR_LEN (EXT4_S_ERR_END - EXT4_S_ERR_START)
+
 #ifdef __KERNEL__
 
 /*
@@ -1143,6 +1169,9 @@ struct ext4_sb_info {
 
 	/* workqueue for dio unwritten */
 	struct workqueue_struct *dio_unwritten_wq;
+
+	/* timer for periodic error stats printing */
+	struct timer_list s_err_report;
 };
 
 static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb)
@@ -1313,6 +1342,10 @@ EXT4_INODE_BIT_FNS(state, state_flags)
 #define EXT4_DEFM_JMODE_DATA	0x0020
 #define EXT4_DEFM_JMODE_ORDERED	0x0040
 #define EXT4_DEFM_JMODE_WBACK	0x0060
+#define EXT4_DEFM_NOBARRIER	0x0100
+#define EXT4_DEFM_BLOCK_VALIDITY 0x0200
+#define EXT4_DEFM_DISCARD	0x0400
+#define EXT4_DEFM_NODELALLOC	0x0800
 
 /*
  * Default journal batch times
@@ -1379,6 +1412,43 @@ struct ext4_dir_entry_2 {
 #define EXT4_MAX_REC_LEN		((1<<16)-1)
 
 /*
+ * If we ever get support for fs block sizes > page_size, we'll need
+ * to remove the #if statements in the next two functions...
+ */
+static inline unsigned int
+ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize)
+{
+	unsigned len = le16_to_cpu(dlen);
+
+#if (PAGE_CACHE_SIZE >= 65536)
+	if (len == EXT4_MAX_REC_LEN || len == 0)
+		return blocksize;
+	return (len & 65532) | ((len & 3) << 16);
+#else
+	return len;
+#endif
+}
+
+static inline __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize)
+{
+	if ((len > blocksize) || (blocksize > (1 << 18)) || (len & 3))
+		BUG();
+#if (PAGE_CACHE_SIZE >= 65536)
+	if (len < 65536)
+		return cpu_to_le16(len);
+	if (len == blocksize) {
+		if (blocksize == 65536)
+			return cpu_to_le16(EXT4_MAX_REC_LEN);
+		else
+			return cpu_to_le16(0);
+	}
+	return cpu_to_le16((len & 65532) | ((len >> 16) & 3));
+#else
+	return cpu_to_le16(len);
+#endif
+}
+
+/*
  * Hash Tree Directory indexing
  * (c) Daniel Phillips, 2001
  */
@@ -1510,9 +1580,11 @@ extern unsigned ext4_init_block_bitmap(struct super_block *sb,
 		ext4_init_block_bitmap(sb, NULL, group, desc)
 
 /* dir.c */
-extern int ext4_check_dir_entry(const char *, struct inode *,
-				struct ext4_dir_entry_2 *,
-				struct buffer_head *, unsigned int);
+extern int __ext4_check_dir_entry(const char *, unsigned int, struct inode *,
+				  struct ext4_dir_entry_2 *,
+				  struct buffer_head *, unsigned int);
+#define ext4_check_dir_entry(dir, de, bh, offset) \
+	__ext4_check_dir_entry(__func__, __LINE__, (dir), (de), (bh), (offset))
 extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash,
 				    __u32 minor_hash,
 				    struct ext4_dir_entry_2 *dirent);
@@ -1601,8 +1673,6 @@ extern long ext4_compat_ioctl(struct file *, unsigned int, unsigned long);
 extern int ext4_ext_migrate(struct inode *);
 
 /* namei.c */
-extern unsigned int ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize);
-extern __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize);
 extern int ext4_orphan_add(handle_t *, struct inode *);
 extern int ext4_orphan_del(handle_t *, struct inode *);
 extern int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash,
@@ -1616,25 +1686,38 @@ extern int ext4_group_extend(struct super_block *sb,
 				ext4_fsblk_t n_blocks_count);
 
 /* super.c */
-extern void __ext4_error(struct super_block *, const char *, const char *, ...)
-	__attribute__ ((format (printf, 3, 4)));
-#define ext4_error(sb, message...)	__ext4_error(sb, __func__, ## message)
-extern void ext4_error_inode(const char *, struct inode *, const char *, ...)
-	__attribute__ ((format (printf, 3, 4)));
-extern void ext4_error_file(const char *, struct file *, const char *, ...)
-	__attribute__ ((format (printf, 3, 4)));
-extern void __ext4_std_error(struct super_block *, const char *, int);
-extern void ext4_abort(struct super_block *, const char *, const char *, ...)
-	__attribute__ ((format (printf, 3, 4)));
-extern void __ext4_warning(struct super_block *, const char *,
+extern void __ext4_error(struct super_block *, const char *, unsigned int,
+			 const char *, ...)
+	__attribute__ ((format (printf, 4, 5)));
+#define ext4_error(sb, message...)	__ext4_error(sb, __func__,	\
+						     __LINE__, ## message)
+extern void ext4_error_inode(struct inode *, const char *, unsigned int,
+			     ext4_fsblk_t, const char *, ...)
+	__attribute__ ((format (printf, 5, 6)));
+extern void ext4_error_file(struct file *, const char *, unsigned int,
+			    const char *, ...)
+	__attribute__ ((format (printf, 4, 5)));
+extern void __ext4_std_error(struct super_block *, const char *,
+			     unsigned int, int);
+extern void __ext4_abort(struct super_block *, const char *, unsigned int,
+		       const char *, ...)
+	__attribute__ ((format (printf, 4, 5)));
+#define ext4_abort(sb, message...)	__ext4_abort(sb, __func__, \
+						       __LINE__, ## message)
+extern void __ext4_warning(struct super_block *, const char *, unsigned int,
 			  const char *, ...)
-	__attribute__ ((format (printf, 3, 4)));
-#define ext4_warning(sb, message...)	__ext4_warning(sb, __func__, ## message)
+	__attribute__ ((format (printf, 4, 5)));
+#define ext4_warning(sb, message...)	__ext4_warning(sb, __func__, \
+						       __LINE__, ## message)
 extern void ext4_msg(struct super_block *, const char *, const char *, ...)
 	__attribute__ ((format (printf, 3, 4)));
-extern void ext4_grp_locked_error(struct super_block *, ext4_group_t,
-				const char *, const char *, ...)
-	__attribute__ ((format (printf, 4, 5)));
+extern void __ext4_grp_locked_error(const char *, unsigned int, \
+				    struct super_block *, ext4_group_t, \
+				    unsigned long, ext4_fsblk_t, \
+				    const char *, ...)
+	__attribute__ ((format (printf, 7, 8)));
+#define ext4_grp_locked_error(sb, grp, message...) \
+	__ext4_grp_locked_error(__func__, __LINE__, (sb), (grp), ## message)
 extern void ext4_update_dynamic_rev(struct super_block *sb);
 extern int ext4_update_compat_feature(handle_t *handle, struct super_block *sb,
 					__u32 compat);
@@ -1768,7 +1851,7 @@ static inline unsigned int ext4_flex_bg_size(struct ext4_sb_info *sbi)
 #define ext4_std_error(sb, errno)				\
 do {								\
 	if ((errno))						\
-		__ext4_std_error((sb), __func__, (errno));	\
+		__ext4_std_error((sb), __func__, __LINE__, (errno));	\
 } while (0)
 
 #ifdef CONFIG_SMP
@@ -1860,6 +1943,12 @@ static inline void ext4_unlock_group(struct super_block *sb,
 	spin_unlock(ext4_group_lock_ptr(sb, group));
 }
 
+static inline void ext4_mark_super_dirty(struct super_block *sb)
+{
+	if (EXT4_SB(sb)->s_journal == NULL)
+		sb->s_dirt =1;
+}
+
 /*
  * Inodes and files operations
  */
@@ -1905,9 +1994,6 @@ extern int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
 			  ssize_t len);
 extern int ext4_map_blocks(handle_t *handle, struct inode *inode,
 			   struct ext4_map_blocks *map, int flags);
-extern int ext4_get_blocks(handle_t *handle, struct inode *inode,
-			   sector_t block, unsigned int max_blocks,
-			   struct buffer_head *bh, int flags);
 extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			__u64 start, __u64 len);
 /* move_extent.c */
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 53d2764d71c..6e272ef6ba9 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -6,29 +6,29 @@
 
 #include <trace/events/ext4.h>
 
-int __ext4_journal_get_undo_access(const char *where, handle_t *handle,
-				struct buffer_head *bh)
+int __ext4_journal_get_undo_access(const char *where, unsigned int line,
+				   handle_t *handle, struct buffer_head *bh)
 {
 	int err = 0;
 
 	if (ext4_handle_valid(handle)) {
 		err = jbd2_journal_get_undo_access(handle, bh);
 		if (err)
-			ext4_journal_abort_handle(where, __func__, bh,
+			ext4_journal_abort_handle(where, line, __func__, bh,
 						  handle, err);
 	}
 	return err;
 }
 
-int __ext4_journal_get_write_access(const char *where, handle_t *handle,
-				struct buffer_head *bh)
+int __ext4_journal_get_write_access(const char *where, unsigned int line,
+				    handle_t *handle, struct buffer_head *bh)
 {
 	int err = 0;
 
 	if (ext4_handle_valid(handle)) {
 		err = jbd2_journal_get_write_access(handle, bh);
 		if (err)
-			ext4_journal_abort_handle(where, __func__, bh,
+			ext4_journal_abort_handle(where, line, __func__, bh,
 						  handle, err);
 	}
 	return err;
@@ -46,9 +46,9 @@ int __ext4_journal_get_write_access(const char *where, handle_t *handle,
  * If the handle isn't valid we're not journaling, but we still need to
  * call into ext4_journal_revoke() to put the buffer head.
  */
-int __ext4_forget(const char *where, handle_t *handle, int is_metadata,
-		  struct inode *inode, struct buffer_head *bh,
-		  ext4_fsblk_t blocknr)
+int __ext4_forget(const char *where, unsigned int line, handle_t *handle,
+		  int is_metadata, struct inode *inode,
+		  struct buffer_head *bh, ext4_fsblk_t blocknr)
 {
 	int err;
 
@@ -79,8 +79,8 @@ int __ext4_forget(const char *where, handle_t *handle, int is_metadata,
 			BUFFER_TRACE(bh, "call jbd2_journal_forget");
 			err = jbd2_journal_forget(handle, bh);
 			if (err)
-				ext4_journal_abort_handle(where, __func__, bh,
-							  handle, err);
+				ext4_journal_abort_handle(where, line, __func__,
+							  bh, handle, err);
 			return err;
 		}
 		return 0;
@@ -92,15 +92,16 @@ int __ext4_forget(const char *where, handle_t *handle, int is_metadata,
 	BUFFER_TRACE(bh, "call jbd2_journal_revoke");
 	err = jbd2_journal_revoke(handle, blocknr, bh);
 	if (err) {
-		ext4_journal_abort_handle(where, __func__, bh, handle, err);
-		ext4_abort(inode->i_sb, __func__,
+		ext4_journal_abort_handle(where, line, __func__,
+					  bh, handle, err);
+		__ext4_abort(inode->i_sb, where, line,
 			   "error %d when attempting revoke", err);
 	}
 	BUFFER_TRACE(bh, "exit");
 	return err;
 }
 
-int __ext4_journal_get_create_access(const char *where,
+int __ext4_journal_get_create_access(const char *where, unsigned int line,
 				handle_t *handle, struct buffer_head *bh)
 {
 	int err = 0;
@@ -108,22 +109,23 @@ int __ext4_journal_get_create_access(const char *where,
 	if (ext4_handle_valid(handle)) {
 		err = jbd2_journal_get_create_access(handle, bh);
 		if (err)
-			ext4_journal_abort_handle(where, __func__, bh,
-						  handle, err);
+			ext4_journal_abort_handle(where, line, __func__,
+						  bh, handle, err);
 	}
 	return err;
 }
 
-int __ext4_handle_dirty_metadata(const char *where, handle_t *handle,
-				 struct inode *inode, struct buffer_head *bh)
+int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
+				 handle_t *handle, struct inode *inode,
+				 struct buffer_head *bh)
 {
 	int err = 0;
 
 	if (ext4_handle_valid(handle)) {
 		err = jbd2_journal_dirty_metadata(handle, bh);
 		if (err)
-			ext4_journal_abort_handle(where, __func__, bh,
-						  handle, err);
+			ext4_journal_abort_handle(where, line, __func__,
+						  bh, handle, err);
 	} else {
 		if (inode)
 			mark_buffer_dirty_inode(bh, inode);
@@ -132,14 +134,33 @@ int __ext4_handle_dirty_metadata(const char *where, handle_t *handle,
 		if (inode && inode_needs_sync(inode)) {
 			sync_dirty_buffer(bh);
 			if (buffer_req(bh) && !buffer_uptodate(bh)) {
-				ext4_error(inode->i_sb,
-					   "IO error syncing inode, "
-					   "inode=%lu, block=%llu",
-					   inode->i_ino,
-					   (unsigned long long) bh->b_blocknr);
+				struct ext4_super_block *es;
+
+				es = EXT4_SB(inode->i_sb)->s_es;
+				es->s_last_error_block =
+					cpu_to_le64(bh->b_blocknr);
+				ext4_error_inode(inode, where, line,
+						 bh->b_blocknr,
+					"IO error syncing itable block");
 				err = -EIO;
 			}
 		}
 	}
 	return err;
 }
+
+int __ext4_handle_dirty_super(const char *where, unsigned int line,
+			      handle_t *handle, struct super_block *sb)
+{
+	struct buffer_head *bh = EXT4_SB(sb)->s_sbh;
+	int err = 0;
+
+	if (ext4_handle_valid(handle)) {
+		err = jbd2_journal_dirty_metadata(handle, bh);
+		if (err)
+			ext4_journal_abort_handle(where, line, __func__,
+						  bh, handle, err);
+	} else
+		sb->s_dirt = 1;
+	return err;
+}
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index dade0c02479..b0bd792c58c 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -122,39 +122,47 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode);
 /*
  * Wrapper functions with which ext4 calls into JBD.
  */
-void ext4_journal_abort_handle(const char *caller, const char *err_fn,
+void ext4_journal_abort_handle(const char *caller, unsigned int line,
+			       const char *err_fn,
 		struct buffer_head *bh, handle_t *handle, int err);
 
-int __ext4_journal_get_undo_access(const char *where, handle_t *handle,
-				struct buffer_head *bh);
+int __ext4_journal_get_undo_access(const char *where, unsigned int line,
+				   handle_t *handle, struct buffer_head *bh);
 
-int __ext4_journal_get_write_access(const char *where, handle_t *handle,
-				struct buffer_head *bh);
+int __ext4_journal_get_write_access(const char *where, unsigned int line,
+				    handle_t *handle, struct buffer_head *bh);
 
-int __ext4_forget(const char *where, handle_t *handle, int is_metadata,
-		  struct inode *inode, struct buffer_head *bh,
-		  ext4_fsblk_t blocknr);
+int __ext4_forget(const char *where, unsigned int line, handle_t *handle,
+		  int is_metadata, struct inode *inode,
+		  struct buffer_head *bh, ext4_fsblk_t blocknr);
 
-int __ext4_journal_get_create_access(const char *where,
+int __ext4_journal_get_create_access(const char *where, unsigned int line,
 				handle_t *handle, struct buffer_head *bh);
 
-int __ext4_handle_dirty_metadata(const char *where, handle_t *handle,
-				 struct inode *inode, struct buffer_head *bh);
+int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
+				 handle_t *handle, struct inode *inode,
+				 struct buffer_head *bh);
+
+int __ext4_handle_dirty_super(const char *where, unsigned int line,
+			      handle_t *handle, struct super_block *sb);
 
 #define ext4_journal_get_undo_access(handle, bh) \
-	__ext4_journal_get_undo_access(__func__, (handle), (bh))
+	__ext4_journal_get_undo_access(__func__, __LINE__, (handle), (bh))
 #define ext4_journal_get_write_access(handle, bh) \
-	__ext4_journal_get_write_access(__func__, (handle), (bh))
+	__ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh))
 #define ext4_forget(handle, is_metadata, inode, bh, block_nr) \
-	__ext4_forget(__func__, (handle), (is_metadata), (inode), (bh),\
-		      (block_nr))
+	__ext4_forget(__func__, __LINE__, (handle), (is_metadata), (inode), \
+		      (bh), (block_nr))
 #define ext4_journal_get_create_access(handle, bh) \
-	__ext4_journal_get_create_access(__func__, (handle), (bh))
+	__ext4_journal_get_create_access(__func__, __LINE__, (handle), (bh))
 #define ext4_handle_dirty_metadata(handle, inode, bh) \
-	__ext4_handle_dirty_metadata(__func__, (handle), (inode), (bh))
+	__ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \
+				     (bh))
+#define ext4_handle_dirty_super(handle, sb) \
+	__ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb))
 
 handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks);
-int __ext4_journal_stop(const char *where, handle_t *handle);
+int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle);
 
 #define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096)
 
@@ -207,7 +215,7 @@ static inline handle_t *ext4_journal_start(struct inode *inode, int nblocks)
 }
 
 #define ext4_journal_stop(handle) \
-	__ext4_journal_stop(__func__, (handle))
+	__ext4_journal_stop(__func__, __LINE__, (handle))
 
 static inline handle_t *ext4_journal_current_handle(void)
 {
@@ -308,17 +316,15 @@ static inline int ext4_should_writeback_data(struct inode *inode)
  * This function controls whether or not we should try to go down the
  * dioread_nolock code paths, which makes it safe to avoid taking
  * i_mutex for direct I/O reads.  This only works for extent-based
- * files, and it doesn't work for nobh or if data journaling is
- * enabled, since the dioread_nolock code uses b_private to pass
- * information back to the I/O completion handler, and this conflicts
- * with the jbd's use of b_private.
+ * files, and it doesn't work if data journaling is enabled, since the
+ * dioread_nolock code uses b_private to pass information back to the
+ * I/O completion handler, and this conflicts with the jbd's use of
+ * b_private.
  */
 static inline int ext4_should_dioread_nolock(struct inode *inode)
 {
 	if (!test_opt(inode->i_sb, DIOREAD_NOLOCK))
 		return 0;
-	if (test_opt(inode->i_sb, NOBH))
-		return 0;
 	if (!S_ISREG(inode->i_mode))
 		return 0;
 	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index bf029c7d551..06328d3e571 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -401,9 +401,9 @@ static int ext4_valid_extent_entries(struct inode *inode,
 	return 1;
 }
 
-static int __ext4_ext_check(const char *function, struct inode *inode,
-					struct ext4_extent_header *eh,
-					int depth)
+static int __ext4_ext_check(const char *function, unsigned int line,
+			    struct inode *inode, struct ext4_extent_header *eh,
+			    int depth)
 {
 	const char *error_msg;
 	int max = 0;
@@ -436,7 +436,7 @@ static int __ext4_ext_check(const char *function, struct inode *inode,
 	return 0;
 
 corrupted:
-	ext4_error_inode(function, inode,
+	ext4_error_inode(inode, function, line, 0,
 			"bad header/extent: %s - magic %x, "
 			"entries %u, max %u(%u), depth %u(%u)",
 			error_msg, le16_to_cpu(eh->eh_magic),
@@ -447,7 +447,7 @@ corrupted:
 }
 
 #define ext4_ext_check(inode, eh, depth)	\
-	__ext4_ext_check(__func__, inode, eh, depth)
+	__ext4_ext_check(__func__, __LINE__, inode, eh, depth)
 
 int ext4_ext_check_inode(struct inode *inode)
 {
@@ -1083,7 +1083,6 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
 {
 	struct ext4_ext_path *curp = path;
 	struct ext4_extent_header *neh;
-	struct ext4_extent_idx *fidx;
 	struct buffer_head *bh;
 	ext4_fsblk_t newblock;
 	int err = 0;
@@ -1144,10 +1143,10 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
 	ext4_idx_store_pblock(curp->p_idx, newblock);
 
 	neh = ext_inode_hdr(inode);
-	fidx = EXT_FIRST_INDEX(neh);
 	ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
 		  le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
-		  le32_to_cpu(fidx->ei_block), idx_pblock(fidx));
+		  le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
+		  idx_pblock(EXT_FIRST_INDEX(neh)));
 
 	neh->eh_depth = cpu_to_le16(path->p_depth + 1);
 	err = ext4_ext_dirty(handle, inode, curp);
@@ -2954,7 +2953,6 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 	struct ext4_extent *ex1 = NULL;
 	struct ext4_extent *ex2 = NULL;
 	struct ext4_extent *ex3 = NULL;
-	struct ext4_extent_header *eh;
 	ext4_lblk_t ee_block, eof_block;
 	unsigned int allocated, ee_len, depth;
 	ext4_fsblk_t newblock;
@@ -2971,7 +2969,6 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 		eof_block = map->m_lblk + map->m_len;
 
 	depth = ext_depth(inode);
-	eh = path[depth].p_hdr;
 	ex = path[depth].p_ext;
 	ee_block = le32_to_cpu(ex->ee_block);
 	ee_len = ext4_ext_get_actual_len(ex);
@@ -3058,7 +3055,6 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 			err = PTR_ERR(path);
 			goto out;
 		}
-		eh = path[depth].p_hdr;
 		ex = path[depth].p_ext;
 		if (ex2 != &newex)
 			ex2 = ex;
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 5313ae4cda2..ee92b66d455 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -70,7 +70,8 @@ ext4_file_write(struct kiocb *iocb, const struct iovec *iov,
 		struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 		size_t length = iov_length(iov, nr_segs);
 
-		if (pos > sbi->s_bitmap_maxbytes)
+		if ((pos > sbi->s_bitmap_maxbytes ||
+		    (pos == sbi->s_bitmap_maxbytes && length > 0)))
 			return -EFBIG;
 
 		if (pos + length > sbi->s_bitmap_maxbytes) {
@@ -123,7 +124,7 @@ static int ext4_file_open(struct inode * inode, struct file * filp)
 		if (!IS_ERR(cp)) {
 			memcpy(sbi->s_es->s_last_mounted, cp,
 			       sizeof(sbi->s_es->s_last_mounted));
-			sb->s_dirt = 1;
+			ext4_mark_super_dirty(sb);
 		}
 	}
 	return dquot_file_open(inode, filp);
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 25c4b3173fd..ac377505ed5 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -279,7 +279,7 @@ out:
 		err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh);
 		if (!fatal)
 			fatal = err;
-		sb->s_dirt = 1;
+		ext4_mark_super_dirty(sb);
 	} else
 		ext4_error(sb, "bit already cleared for inode %lu", ino);
 
@@ -965,7 +965,7 @@ got:
 	percpu_counter_dec(&sbi->s_freeinodes_counter);
 	if (S_ISDIR(mode))
 		percpu_counter_inc(&sbi->s_dirs_counter);
-	sb->s_dirt = 1;
+	ext4_mark_super_dirty(sb);
 
 	if (sbi->s_log_groups_per_flex) {
 		flex_group = ext4_flex_group(sbi, group);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 0afc8c1d8cf..a0ab3754d0d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -221,6 +221,7 @@ void ext4_delete_inode(struct inode *inode)
 				     "couldn't extend journal (err %d)", err);
 		stop_handle:
 			ext4_journal_stop(handle);
+			ext4_orphan_del(NULL, inode);
 			goto no_delete;
 		}
 	}
@@ -337,9 +338,11 @@ static int ext4_block_to_path(struct inode *inode,
 	return n;
 }
 
-static int __ext4_check_blockref(const char *function, struct inode *inode,
+static int __ext4_check_blockref(const char *function, unsigned int line,
+				 struct inode *inode,
 				 __le32 *p, unsigned int max)
 {
+	struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
 	__le32 *bref = p;
 	unsigned int blk;
 
@@ -348,8 +351,9 @@ static int __ext4_check_blockref(const char *function, struct inode *inode,
 		if (blk &&
 		    unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb),
 						    blk, 1))) {
-			ext4_error_inode(function, inode,
-					 "invalid block reference %u", blk);
+			es->s_last_error_block = cpu_to_le64(blk);
+			ext4_error_inode(inode, function, line, blk,
+					 "invalid block");
 			return -EIO;
 		}
 	}
@@ -358,11 +362,13 @@ static int __ext4_check_blockref(const char *function, struct inode *inode,
 
 
 #define ext4_check_indirect_blockref(inode, bh)                         \
-	__ext4_check_blockref(__func__, inode, (__le32 *)(bh)->b_data,  \
+	__ext4_check_blockref(__func__, __LINE__, inode,		\
+			      (__le32 *)(bh)->b_data,			\
 			      EXT4_ADDR_PER_BLOCK((inode)->i_sb))
 
 #define ext4_check_inode_blockref(inode)                                \
-	__ext4_check_blockref(__func__, inode, EXT4_I(inode)->i_data,   \
+	__ext4_check_blockref(__func__, __LINE__, inode,		\
+			      EXT4_I(inode)->i_data,			\
 			      EXT4_NDIR_BLOCKS)
 
 /**
@@ -1128,20 +1134,24 @@ void ext4_da_update_reserve_space(struct inode *inode,
 		ext4_discard_preallocations(inode);
 }
 
-static int check_block_validity(struct inode *inode, const char *func,
+static int __check_block_validity(struct inode *inode, const char *func,
+				unsigned int line,
 				struct ext4_map_blocks *map)
 {
 	if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), map->m_pblk,
 				   map->m_len)) {
-		ext4_error_inode(func, inode,
-			   "lblock %lu mapped to illegal pblock %llu "
-			   "(length %d)", (unsigned long) map->m_lblk,
-				 map->m_pblk, map->m_len);
+		ext4_error_inode(inode, func, line, map->m_pblk,
+				 "lblock %lu mapped to illegal pblock "
+				 "(length %d)", (unsigned long) map->m_lblk,
+				 map->m_len);
 		return -EIO;
 	}
 	return 0;
 }
 
+#define check_block_validity(inode, map)	\
+	__check_block_validity((inode), __func__, __LINE__, (map))
+
 /*
  * Return the number of contiguous dirty pages in a given inode
  * starting at page frame idx.
@@ -1244,7 +1254,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
 	up_read((&EXT4_I(inode)->i_data_sem));
 
 	if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) {
-		int ret = check_block_validity(inode, __func__, map);
+		int ret = check_block_validity(inode, map);
 		if (ret != 0)
 			return ret;
 	}
@@ -1324,9 +1334,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
 
 	up_write((&EXT4_I(inode)->i_data_sem));
 	if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) {
-		int ret = check_block_validity(inode,
-					       "ext4_map_blocks_after_alloc",
-					       map);
+		int ret = check_block_validity(inode, map);
 		if (ret != 0)
 			return ret;
 	}
@@ -1519,9 +1527,25 @@ static int walk_page_buffers(handle_t *handle,
 static int do_journal_get_write_access(handle_t *handle,
 				       struct buffer_head *bh)
 {
+	int dirty = buffer_dirty(bh);
+	int ret;
+
 	if (!buffer_mapped(bh) || buffer_freed(bh))
 		return 0;
-	return ext4_journal_get_write_access(handle, bh);
+	/*
+	 * __block_prepare_write() could have dirtied some buffers. Clean
+	 * the dirty bit as jbd2_journal_get_write_access() could complain
+	 * otherwise about fs integrity issues. Setting of the dirty bit
+	 * by __block_prepare_write() isn't a real problem here as we clear
+	 * the bit before releasing a page lock and thus writeback cannot
+	 * ever write the buffer.
+	 */
+	if (dirty)
+		clear_buffer_dirty(bh);
+	ret = ext4_journal_get_write_access(handle, bh);
+	if (!ret && dirty)
+		ret = ext4_handle_dirty_metadata(handle, NULL, bh);
+	return ret;
 }
 
 /*
@@ -2194,7 +2218,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 	BUG_ON(!handle);
 
 	/*
-	 * Call ext4_get_blocks() to allocate any delayed allocation
+	 * Call ext4_map_blocks() to allocate any delayed allocation
 	 * blocks, or to convert an uninitialized extent to be
 	 * initialized (in the case where we have written into
 	 * one or more preallocated blocks).
@@ -2203,7 +2227,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 	 * indicate that we are on the delayed allocation path.  This
 	 * affects functions in many different parts of the allocation
 	 * call path.  This flag exists primarily because we don't
-	 * want to change *many* call functions, so ext4_get_blocks()
+	 * want to change *many* call functions, so ext4_map_blocks()
 	 * will set the magic i_delalloc_reserved_flag once the
 	 * inode's allocation semaphore is taken.
 	 *
@@ -2221,6 +2245,8 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 
 	blks = ext4_map_blocks(handle, mpd->inode, &map, get_blocks_flags);
 	if (blks < 0) {
+		struct super_block *sb = mpd->inode->i_sb;
+
 		err = blks;
 		/*
 		 * If get block returns with error we simply
@@ -2231,7 +2257,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 			return 0;
 
 		if (err == -ENOSPC &&
-		    ext4_count_free_blocks(mpd->inode->i_sb)) {
+		    ext4_count_free_blocks(sb)) {
 			mpd->retval = err;
 			return 0;
 		}
@@ -2243,16 +2269,17 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 		 * writepage and writepages will again try to write
 		 * the same.
 		 */
-		ext4_msg(mpd->inode->i_sb, KERN_CRIT,
-			 "delayed block allocation failed for inode %lu at "
-			 "logical offset %llu with max blocks %zd with "
-			 "error %d", mpd->inode->i_ino,
-			 (unsigned long long) next,
-			 mpd->b_size >> mpd->inode->i_blkbits, err);
-		printk(KERN_CRIT "This should not happen!!  "
-		       "Data will be lost\n");
-		if (err == -ENOSPC) {
-			ext4_print_free_blocks(mpd->inode);
+		if (!(EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED)) {
+			ext4_msg(sb, KERN_CRIT,
+				 "delayed block allocation failed for inode %lu "
+				 "at logical offset %llu with max blocks %zd "
+				 "with error %d", mpd->inode->i_ino,
+				 (unsigned long long) next,
+				 mpd->b_size >> mpd->inode->i_blkbits, err);
+			ext4_msg(sb, KERN_CRIT,
+				"This should not happen!! Data will be lost\n");
+			if (err == -ENOSPC)
+				ext4_print_free_blocks(mpd->inode);
 		}
 		/* invalidate all the pages */
 		ext4_da_block_invalidatepages(mpd, next,
@@ -2320,7 +2347,7 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd,
 	 * XXX Don't go larger than mballoc is willing to allocate
 	 * This is a stopgap solution.  We eventually need to fold
 	 * mpage_da_submit_io() into this function and then call
-	 * ext4_get_blocks() multiple times in a loop
+	 * ext4_map_blocks() multiple times in a loop
 	 */
 	if (nrblocks >= 8*1024*1024/mpd->inode->i_sb->s_blocksize)
 		goto flush_it;
@@ -2553,18 +2580,16 @@ static int ext4_da_get_block_prep(struct inode *inode, sector_t iblock,
 /*
  * This function is used as a standard get_block_t calback function
  * when there is no desire to allocate any blocks.  It is used as a
- * callback function for block_prepare_write(), nobh_writepage(), and
- * block_write_full_page().  These functions should only try to map a
- * single block at a time.
+ * callback function for block_prepare_write() and block_write_full_page().
+ * These functions should only try to map a single block at a time.
  *
  * Since this function doesn't do block allocations even if the caller
  * requests it by passing in create=1, it is critically important that
  * any caller checks to make sure that any buffer heads are returned
  * by this function are either all already mapped or marked for
- * delayed allocation before calling nobh_writepage() or
- * block_write_full_page().  Otherwise, b_blocknr could be left
- * unitialized, and the page write functions will be taken by
- * surprise.
+ * delayed allocation before calling  block_write_full_page().  Otherwise,
+ * b_blocknr could be left unitialized, and the page write functions will
+ * be taken by surprise.
  */
 static int noalloc_get_block_write(struct inode *inode, sector_t iblock,
 				   struct buffer_head *bh_result, int create)
@@ -2749,9 +2774,7 @@ static int ext4_writepage(struct page *page,
 		return __ext4_journalled_writepage(page, len);
 	}
 
-	if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode))
-		ret = nobh_writepage(page, noalloc_get_block_write, wbc);
-	else if (page_bufs && buffer_uninit(page_bufs)) {
+	if (page_bufs && buffer_uninit(page_bufs)) {
 		ext4_set_bh_endio(page_bufs, inode);
 		ret = block_write_full_page_endio(page, noalloc_get_block_write,
 					    wbc, ext4_end_io_buffer_write);
@@ -3146,13 +3169,10 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	int ret, retries = 0;
 	struct page *page;
 	pgoff_t index;
-	unsigned from, to;
 	struct inode *inode = mapping->host;
 	handle_t *handle;
 
 	index = pos >> PAGE_CACHE_SHIFT;
-	from = pos & (PAGE_CACHE_SIZE - 1);
-	to = from + len;
 
 	if (ext4_nonda_switch(inode->i_sb)) {
 		*fsdata = (void *)FALL_BACK_TO_NONDELALLOC;
@@ -3668,6 +3688,8 @@ static int ext4_end_io_nolock(ext4_io_end_t *io)
 		return ret;
 	}
 
+	if (io->iocb)
+		aio_complete(io->iocb, io->result, 0);
 	/* clear the DIO AIO unwritten flag */
 	io->flag = 0;
 	return ret;
@@ -3767,6 +3789,8 @@ static ext4_io_end_t *ext4_init_io_end (struct inode *inode, gfp_t flags)
 		io->offset = 0;
 		io->size = 0;
 		io->page = NULL;
+		io->iocb = NULL;
+		io->result = 0;
 		INIT_WORK(&io->work, ext4_end_io_work);
 		INIT_LIST_HEAD(&io->list);
 	}
@@ -3796,12 +3820,18 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 	if (io_end->flag != EXT4_IO_UNWRITTEN){
 		ext4_free_io_end(io_end);
 		iocb->private = NULL;
-		goto out;
+out:
+		if (is_async)
+			aio_complete(iocb, ret, 0);
+		return;
 	}
 
 	io_end->offset = offset;
 	io_end->size = size;
-	io_end->flag = EXT4_IO_UNWRITTEN;
+	if (is_async) {
+		io_end->iocb = iocb;
+		io_end->result = ret;
+	}
 	wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq;
 
 	/* queue the work to convert unwritten extents to written */
@@ -3813,9 +3843,6 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 	list_add_tail(&io_end->list, &ei->i_completed_io_list);
 	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
 	iocb->private = NULL;
-out:
-	if (is_async)
-		aio_complete(iocb, ret, 0);
 }
 
 static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
@@ -3941,7 +3968,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 				return -ENOMEM;
 			/*
 			 * we save the io structure for current async
-			 * direct IO, so that later ext4_get_blocks()
+			 * direct IO, so that later ext4_map_blocks()
 			 * could flag the io structure whether there
 			 * is a unwritten extents needs to be converted
 			 * when IO is completed.
@@ -4132,17 +4159,6 @@ int ext4_block_truncate_page(handle_t *handle,
 	length = blocksize - (offset & (blocksize - 1));
 	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);
 
-	/*
-	 * For "nobh" option,  we can only work if we don't need to
-	 * read-in the page - otherwise we create buffers to do the IO.
-	 */
-	if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) &&
-	     ext4_should_writeback_data(inode) && PageUptodate(page)) {
-		zero_user(page, offset, length);
-		set_page_dirty(page);
-		goto unlock;
-	}
-
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, blocksize, 0);
 
@@ -4492,9 +4508,8 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode,
 			 * (should be rare).
 			 */
 			if (!bh) {
-				EXT4_ERROR_INODE(inode,
-						 "Read failure block=%llu",
-						 (unsigned long long) nr);
+				EXT4_ERROR_INODE_BLOCK(inode, nr,
+						       "Read failure");
 				continue;
 			}
 
@@ -4506,27 +4521,6 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode,
 					depth);
 
 			/*
-			 * We've probably journalled the indirect block several
-			 * times during the truncate.  But it's no longer
-			 * needed and we now drop it from the transaction via
-			 * jbd2_journal_revoke().
-			 *
-			 * That's easy if it's exclusively part of this
-			 * transaction.  But if it's part of the committing
-			 * transaction then jbd2_journal_forget() will simply
-			 * brelse() it.  That means that if the underlying
-			 * block is reallocated in ext4_get_block(),
-			 * unmap_underlying_metadata() will find this block
-			 * and will try to get rid of it.  damn, damn.
-			 *
-			 * If this block has already been committed to the
-			 * journal, a revoke record will be written.  And
-			 * revoke records must be emitted *before* clearing
-			 * this block's bit in the bitmaps.
-			 */
-			ext4_forget(handle, 1, inode, bh, bh->b_blocknr);
-
-			/*
 			 * Everything below this this pointer has been
 			 * released.  Now let this top-of-subtree go.
 			 *
@@ -4550,8 +4544,20 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode,
 					    blocks_for_truncate(inode));
 			}
 
+			/*
+			 * The forget flag here is critical because if
+			 * we are journaling (and not doing data
+			 * journaling), we have to make sure a revoke
+			 * record is written to prevent the journal
+			 * replay from overwriting the (former)
+			 * indirect block if it gets reallocated as a
+			 * data block.  This must happen in the same
+			 * transaction where the data blocks are
+			 * actually freed.
+			 */
 			ext4_free_blocks(handle, inode, 0, nr, 1,
-					 EXT4_FREE_BLOCKS_METADATA);
+					 EXT4_FREE_BLOCKS_METADATA|
+					 EXT4_FREE_BLOCKS_FORGET);
 
 			if (parent_bh) {
 				/*
@@ -4809,8 +4815,8 @@ static int __ext4_get_inode_loc(struct inode *inode,
 
 	bh = sb_getblk(sb, block);
 	if (!bh) {
-		EXT4_ERROR_INODE(inode, "unable to read inode block - "
-				 "block %llu", block);
+		EXT4_ERROR_INODE_BLOCK(inode, block,
+				       "unable to read itable block");
 		return -EIO;
 	}
 	if (!buffer_uptodate(bh)) {
@@ -4908,8 +4914,8 @@ make_io:
 		submit_bh(READ_META, bh);
 		wait_on_buffer(bh);
 		if (!buffer_uptodate(bh)) {
-			EXT4_ERROR_INODE(inode, "unable to read inode "
-					 "block %llu", block);
+			EXT4_ERROR_INODE_BLOCK(inode, block,
+					       "unable to read itable block");
 			brelse(bh);
 			return -EIO;
 		}
@@ -4980,7 +4986,7 @@ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode,
 		/* we are using combined 48 bit field */
 		i_blocks = ((u64)le16_to_cpu(raw_inode->i_blocks_high)) << 32 |
 					le32_to_cpu(raw_inode->i_blocks_lo);
-		if (ei->i_flags & EXT4_HUGE_FILE_FL) {
+		if (ext4_test_inode_flag(inode, EXT4_INODE_HUGE_FILE)) {
 			/* i_blocks represent file system block size */
 			return i_blocks  << (inode->i_blkbits - 9);
 		} else {
@@ -5076,7 +5082,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 		transaction_t *transaction;
 		tid_t tid;
 
-		spin_lock(&journal->j_state_lock);
+		read_lock(&journal->j_state_lock);
 		if (journal->j_running_transaction)
 			transaction = journal->j_running_transaction;
 		else
@@ -5085,7 +5091,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 			tid = transaction->t_tid;
 		else
 			tid = journal->j_commit_sequence;
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		ei->i_sync_tid = tid;
 		ei->i_datasync_tid = tid;
 	}
@@ -5130,7 +5136,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 				 ei->i_file_acl);
 		ret = -EIO;
 		goto bad_inode;
-	} else if (ei->i_flags & EXT4_EXTENTS_FL) {
+	} else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) ||
 		    (S_ISLNK(inode->i_mode) &&
 		     !ext4_inode_is_fast_symlink(inode)))
@@ -5410,9 +5416,8 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 		if (wbc->sync_mode == WB_SYNC_ALL)
 			sync_dirty_buffer(iloc.bh);
 		if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) {
-			EXT4_ERROR_INODE(inode,
-				"IO error syncing inode (block=%llu)",
-				(unsigned long long) iloc.bh->b_blocknr);
+			EXT4_ERROR_INODE_BLOCK(inode, iloc.bh->b_blocknr,
+					 "IO error syncing inode");
 			err = -EIO;
 		}
 		brelse(iloc.bh);
@@ -5487,10 +5492,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 		if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 			struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 
-			if (attr->ia_size > sbi->s_bitmap_maxbytes) {
-				error = -EFBIG;
-				goto err_out;
-			}
+			if (attr->ia_size > sbi->s_bitmap_maxbytes)
+				return -EFBIG;
 		}
 	}
 
@@ -5692,7 +5695,7 @@ int ext4_writepage_trans_blocks(struct inode *inode)
  * Calculate the journal credits for a chunk of data modification.
  *
  * This is called from DIO, fallocate or whoever calling
- * ext4_get_blocks() to map/allocate a chunk of contiguous disk blocks.
+ * ext4_map_blocks() to map/allocate a chunk of contiguous disk blocks.
  *
  * journal buffers for data blocks are not included here, as DIO
  * and fallocate do no need to journal data buffers.
@@ -5758,7 +5761,6 @@ static int ext4_expand_extra_isize(struct inode *inode,
 {
 	struct ext4_inode *raw_inode;
 	struct ext4_xattr_ibody_header *header;
-	struct ext4_xattr_entry *entry;
 
 	if (EXT4_I(inode)->i_extra_isize >= new_extra_isize)
 		return 0;
@@ -5766,7 +5768,6 @@ static int ext4_expand_extra_isize(struct inode *inode,
 	raw_inode = ext4_raw_inode(&iloc);
 
 	header = IHDR(inode, raw_inode);
-	entry = IFIRST(header);
 
 	/* No extended attributes present */
 	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR) ||
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 0e83dfd351d..4b4ad4b7ce5 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -446,10 +446,11 @@ static void mb_free_blocks_double(struct inode *inode, struct ext4_buddy *e4b,
 			blocknr = ext4_group_first_block_no(sb, e4b->bd_group);
 			blocknr += first + i;
 			ext4_grp_locked_error(sb, e4b->bd_group,
-				   __func__, "double-free of inode"
-				   " %lu's block %llu(bit %u in group %u)",
-				   inode ? inode->i_ino : 0, blocknr,
-				   first + i, e4b->bd_group);
+					      inode ? inode->i_ino : 0,
+					      blocknr,
+					      "freeing block already freed "
+					      "(bit %u)",
+					      first + i);
 		}
 		mb_clear_bit(first + i, e4b->bd_info->bb_bitmap);
 	}
@@ -712,9 +713,9 @@ void ext4_mb_generate_buddy(struct super_block *sb,
 	grp->bb_fragments = fragments;
 
 	if (free != grp->bb_free) {
-		ext4_grp_locked_error(sb, group,  __func__,
-			"EXT4-fs: group %u: %u blocks in bitmap, %u in gd",
-			group, free, grp->bb_free);
+		ext4_grp_locked_error(sb, group, 0, 0,
+				      "%u blocks in bitmap, %u in gd",
+				      free, grp->bb_free);
 		/*
 		 * If we intent to continue, we consider group descritor
 		 * corrupt and update bb_free using bitmap value
@@ -1296,10 +1297,10 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
 			blocknr = ext4_group_first_block_no(sb, e4b->bd_group);
 			blocknr += block;
 			ext4_grp_locked_error(sb, e4b->bd_group,
-				   __func__, "double-free of inode"
-				   " %lu's block %llu(bit %u in group %u)",
-				   inode ? inode->i_ino : 0, blocknr, block,
-				   e4b->bd_group);
+					      inode ? inode->i_ino : 0,
+					      blocknr,
+					      "freeing already freed block "
+					      "(bit %u)", block);
 		}
 		mb_clear_bit(block, EXT4_MB_BITMAP(e4b));
 		e4b->bd_info->bb_counters[order]++;
@@ -1788,8 +1789,8 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac,
 			 * free blocks even though group info says we
 			 * we have free blocks
 			 */
-			ext4_grp_locked_error(sb, e4b->bd_group,
-					__func__, "%d free blocks as per "
+			ext4_grp_locked_error(sb, e4b->bd_group, 0, 0,
+					"%d free blocks as per "
 					"group info. But bitmap says 0",
 					free);
 			break;
@@ -1798,8 +1799,8 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac,
 		mb_find_extent(e4b, 0, i, ac->ac_g_ex.fe_len, &ex);
 		BUG_ON(ex.fe_len <= 0);
 		if (free < ex.fe_len) {
-			ext4_grp_locked_error(sb, e4b->bd_group,
-					__func__, "%d free blocks as per "
+			ext4_grp_locked_error(sb, e4b->bd_group, 0, 0,
+					"%d free blocks as per "
 					"group info. But got %d blocks",
 					free, ex.fe_len);
 			/*
@@ -1821,8 +1822,7 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac,
 
 /*
  * This is a special case for storages like raid5
- * we try to find stripe-aligned chunks for stripe-size requests
- * XXX should do so at least for multiples of stripe size as well
+ * we try to find stripe-aligned chunks for stripe-size-multiple requests
  */
 static noinline_for_stack
 void ext4_mb_scan_aligned(struct ext4_allocation_context *ac,
@@ -1999,7 +1999,6 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
 	ext4_group_t ngroups, group, i;
 	int cr;
 	int err = 0;
-	int bsbits;
 	struct ext4_sb_info *sbi;
 	struct super_block *sb;
 	struct ext4_buddy e4b;
@@ -2041,8 +2040,6 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
 			ac->ac_2order = i - 1;
 	}
 
-	bsbits = ac->ac_sb->s_blocksize_bits;
-
 	/* if stream allocation is enabled, use global goal */
 	if (ac->ac_flags & EXT4_MB_STREAM_ALLOC) {
 		/* TBD: may be hot point */
@@ -2094,8 +2091,8 @@ repeat:
 			ac->ac_groups_scanned++;
 			if (cr == 0)
 				ext4_mb_simple_scan_group(ac, &e4b);
-			else if (cr == 1 &&
-					ac->ac_g_ex.fe_len == sbi->s_stripe)
+			else if (cr == 1 && sbi->s_stripe &&
+					!(ac->ac_g_ex.fe_len % sbi->s_stripe))
 				ext4_mb_scan_aligned(ac, &e4b);
 			else
 				ext4_mb_complex_scan_group(ac, &e4b);
@@ -2221,7 +2218,7 @@ static int ext4_mb_seq_groups_open(struct inode *inode, struct file *file)
 
 	rc = seq_open(file, &ext4_mb_seq_groups_ops);
 	if (rc == 0) {
-		struct seq_file *m = (struct seq_file *)file->private_data;
+		struct seq_file *m = file->private_data;
 		m->private = sb;
 	}
 	return rc;
@@ -2560,6 +2557,22 @@ int ext4_mb_release(struct super_block *sb)
 	return 0;
 }
 
+static inline void ext4_issue_discard(struct super_block *sb,
+		ext4_group_t block_group, ext4_grpblk_t block, int count)
+{
+	int ret;
+	ext4_fsblk_t discard_block;
+
+	discard_block = block + ext4_group_first_block_no(sb, block_group);
+	trace_ext4_discard_blocks(sb,
+			(unsigned long long) discard_block, count);
+	ret = sb_issue_discard(sb, discard_block, count);
+	if (ret == EOPNOTSUPP) {
+		ext4_warning(sb, "discard not supported, disabling");
+		clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD);
+	}
+}
+
 /*
  * This function is called by the jbd2 layer once the commit has finished,
  * so we know we can free the blocks that were released with that commit.
@@ -2579,22 +2592,9 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
 		mb_debug(1, "gonna free %u blocks in group %u (0x%p):",
 			 entry->count, entry->group, entry);
 
-		if (test_opt(sb, DISCARD)) {
-			int ret;
-			ext4_fsblk_t discard_block;
-
-			discard_block = entry->start_blk +
-				ext4_group_first_block_no(sb, entry->group);
-			trace_ext4_discard_blocks(sb,
-					(unsigned long long)discard_block,
-					entry->count);
-			ret = sb_issue_discard(sb, discard_block, entry->count);
-			if (ret == EOPNOTSUPP) {
-				ext4_warning(sb,
-					"discard not supported, disabling");
-				clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD);
-			}
-		}
+		if (test_opt(sb, DISCARD))
+			ext4_issue_discard(sb, entry->group,
+					entry->start_blk, entry->count);
 
 		err = ext4_mb_load_buddy(sb, entry->group, &e4b);
 		/* we expect to find existing buddy because it's pinned */
@@ -2712,7 +2712,6 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
 				handle_t *handle, unsigned int reserv_blks)
 {
 	struct buffer_head *bitmap_bh = NULL;
-	struct ext4_super_block *es;
 	struct ext4_group_desc *gdp;
 	struct buffer_head *gdp_bh;
 	struct ext4_sb_info *sbi;
@@ -2725,8 +2724,6 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
 
 	sb = ac->ac_sb;
 	sbi = EXT4_SB(sb);
-	es = sbi->s_es;
-
 
 	err = -EIO;
 	bitmap_bh = ext4_read_block_bitmap(sb, ac->ac_b_ex.fe_group);
@@ -2812,7 +2809,7 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
 	err = ext4_handle_dirty_metadata(handle, NULL, gdp_bh);
 
 out_err:
-	sb->s_dirt = 1;
+	ext4_mark_super_dirty(sb);
 	brelse(bitmap_bh);
 	return err;
 }
@@ -2850,7 +2847,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac,
 	int bsbits, max;
 	ext4_lblk_t end;
 	loff_t size, orig_size, start_off;
-	ext4_lblk_t start, orig_start;
+	ext4_lblk_t start;
 	struct ext4_inode_info *ei = EXT4_I(ac->ac_inode);
 	struct ext4_prealloc_space *pa;
 
@@ -2881,6 +2878,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac,
 	size = size << bsbits;
 	if (size < i_size_read(ac->ac_inode))
 		size = i_size_read(ac->ac_inode);
+	orig_size = size;
 
 	/* max size of free chunks */
 	max = 2 << bsbits;
@@ -2922,8 +2920,8 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac,
 		start_off = (loff_t)ac->ac_o_ex.fe_logical << bsbits;
 		size	  = ac->ac_o_ex.fe_len << bsbits;
 	}
-	orig_size = size = size >> bsbits;
-	orig_start = start = start_off >> bsbits;
+	size = size >> bsbits;
+	start = start_off >> bsbits;
 
 	/* don't cover already allocated blocks in selected range */
 	if (ar->pleft && start <= ar->lleft) {
@@ -3547,7 +3545,6 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 	ext4_group_t group;
 	ext4_grpblk_t bit;
 	unsigned long long grp_blk_start;
-	sector_t start;
 	int err = 0;
 	int free = 0;
 
@@ -3567,10 +3564,9 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 		if (bit >= end)
 			break;
 		next = mb_find_next_bit(bitmap_bh->b_data, end, bit);
-		start = ext4_group_first_block_no(sb, group) + bit;
 		mb_debug(1, "    free preallocated %u/%u in group %u\n",
-				(unsigned) start, (unsigned) next - bit,
-				(unsigned) group);
+			 (unsigned) ext4_group_first_block_no(sb, group) + bit,
+			 (unsigned) next - bit, (unsigned) group);
 		free += next - bit;
 
 		if (ac) {
@@ -3581,7 +3577,7 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 			trace_ext4_mballoc_discard(ac);
 		}
 
-		trace_ext4_mb_release_inode_pa(ac, pa, grp_blk_start + bit,
+		trace_ext4_mb_release_inode_pa(sb, ac, pa, grp_blk_start + bit,
 					       next - bit);
 		mb_free_blocks(pa->pa_inode, e4b, bit, next - bit);
 		bit = next + 1;
@@ -3591,8 +3587,7 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 			pa, (unsigned long) pa->pa_lstart,
 			(unsigned long) pa->pa_pstart,
 			(unsigned long) pa->pa_len);
-		ext4_grp_locked_error(sb, group,
-					__func__, "free %u, pa_free %u",
+		ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u",
 					free, pa->pa_free);
 		/*
 		 * pa is already deleted so we use the value obtained
@@ -3613,7 +3608,7 @@ ext4_mb_release_group_pa(struct ext4_buddy *e4b,
 	ext4_group_t group;
 	ext4_grpblk_t bit;
 
-	trace_ext4_mb_release_group_pa(ac, pa);
+	trace_ext4_mb_release_group_pa(sb, ac, pa);
 	BUG_ON(pa->pa_deleted == 0);
 	ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit);
 	BUG_ON(group != e4b->bd_group && pa->pa_len != 0);
@@ -3889,6 +3884,9 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac)
 	struct super_block *sb = ac->ac_sb;
 	ext4_group_t ngroups, i;
 
+	if (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED)
+		return;
+
 	printk(KERN_ERR "EXT4-fs: Can't allocate:"
 			" Allocation context details:\n");
 	printk(KERN_ERR "EXT4-fs: status %d flags %d\n",
@@ -4255,7 +4253,7 @@ static int ext4_mb_discard_preallocations(struct super_block *sb, int needed)
  * to usual allocation
  */
 ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
-				 struct ext4_allocation_request *ar, int *errp)
+				struct ext4_allocation_request *ar, int *errp)
 {
 	int freed;
 	struct ext4_allocation_context *ac = NULL;
@@ -4299,7 +4297,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
 		inquota = ar->len;
 		if (ar->len == 0) {
 			*errp = -EDQUOT;
-			goto out3;
+			goto out;
 		}
 	}
 
@@ -4307,13 +4305,13 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
 	if (!ac) {
 		ar->len = 0;
 		*errp = -ENOMEM;
-		goto out1;
+		goto out;
 	}
 
 	*errp = ext4_mb_initialize_context(ac, ar);
 	if (*errp) {
 		ar->len = 0;
-		goto out2;
+		goto out;
 	}
 
 	ac->ac_op = EXT4_MB_HISTORY_PREALLOC;
@@ -4322,7 +4320,9 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
 		ext4_mb_normalize_request(ac, ar);
 repeat:
 		/* allocate space in core */
-		ext4_mb_regular_allocator(ac);
+		*errp = ext4_mb_regular_allocator(ac);
+		if (*errp)
+			goto errout;
 
 		/* as we've just preallocated more space than
 		 * user requested orinally, we store allocated
@@ -4333,7 +4333,7 @@ repeat:
 	}
 	if (likely(ac->ac_status == AC_STATUS_FOUND)) {
 		*errp = ext4_mb_mark_diskspace_used(ac, handle, reserv_blks);
-		if (*errp ==  -EAGAIN) {
+		if (*errp == -EAGAIN) {
 			/*
 			 * drop the reference that we took
 			 * in ext4_mb_use_best_found
@@ -4344,12 +4344,10 @@ repeat:
 			ac->ac_b_ex.fe_len = 0;
 			ac->ac_status = AC_STATUS_CONTINUE;
 			goto repeat;
-		} else if (*errp) {
+		} else if (*errp)
+		errout:
 			ext4_discard_allocated_blocks(ac);
-			ac->ac_b_ex.fe_len = 0;
-			ar->len = 0;
-			ext4_mb_show_ac(ac);
-		} else {
+		else {
 			block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex);
 			ar->len = ac->ac_b_ex.fe_len;
 		}
@@ -4358,19 +4356,19 @@ repeat:
 		if (freed)
 			goto repeat;
 		*errp = -ENOSPC;
+	}
+
+	if (*errp) {
 		ac->ac_b_ex.fe_len = 0;
 		ar->len = 0;
 		ext4_mb_show_ac(ac);
 	}
-
 	ext4_mb_release_context(ac);
-
-out2:
-	kmem_cache_free(ext4_ac_cachep, ac);
-out1:
+out:
+	if (ac)
+		kmem_cache_free(ext4_ac_cachep, ac);
 	if (inquota && ar->len < inquota)
 		dquot_free_block(ar->inode, inquota - ar->len);
-out3:
 	if (!ar->len) {
 		if (!EXT4_I(ar->inode)->i_delalloc_reserved_flag)
 			/* release all the reserved blocks if non delalloc */
@@ -4402,6 +4400,7 @@ static noinline_for_stack int
 ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 		      struct ext4_free_data *new_entry)
 {
+	ext4_group_t group = e4b->bd_group;
 	ext4_grpblk_t block;
 	struct ext4_free_data *entry;
 	struct ext4_group_info *db = e4b->bd_info;
@@ -4434,9 +4433,9 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 		else if (block >= (entry->start_blk + entry->count))
 			n = &(*n)->rb_right;
 		else {
-			ext4_grp_locked_error(sb, e4b->bd_group, __func__,
-					"Double free of blocks %d (%d %d)",
-					block, entry->start_blk, entry->count);
+			ext4_grp_locked_error(sb, group, 0,
+				ext4_group_first_block_no(sb, group) + block,
+				"Block already on to-be-freed list");
 			return 0;
 		}
 	}
@@ -4494,7 +4493,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
 	struct super_block *sb = inode->i_sb;
 	struct ext4_allocation_context *ac = NULL;
 	struct ext4_group_desc *gdp;
-	struct ext4_super_block *es;
 	unsigned long freed = 0;
 	unsigned int overflow;
 	ext4_grpblk_t bit;
@@ -4513,7 +4511,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
 	}
 
 	sbi = EXT4_SB(sb);
-	es = EXT4_SB(sb)->s_es;
 	if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) &&
 	    !ext4_data_block_valid(sbi, block, count)) {
 		ext4_error(sb, "Freeing blocks not in datazone - "
@@ -4647,6 +4644,8 @@ do_more:
 		mb_clear_bits(bitmap_bh->b_data, bit, count);
 		mb_free_blocks(inode, &e4b, bit, count);
 		ext4_mb_return_to_preallocation(inode, &e4b, block, count);
+		if (test_opt(sb, DISCARD))
+			ext4_issue_discard(sb, block_group, bit, count);
 	}
 
 	ret = ext4_free_blks_count(sb, gdp) + count;
@@ -4680,7 +4679,7 @@ do_more:
 		put_bh(bitmap_bh);
 		goto do_more;
 	}
-	sb->s_dirt = 1;
+	ext4_mark_super_dirty(sb);
 error_return:
 	if (freed)
 		dquot_free_block(inode, freed);
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 6f3a27ec30b..1765c2c50a9 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -376,7 +376,7 @@ static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode,
 	 * We have the extent map build with the tmp inode.
 	 * Now copy the i_data across
 	 */
-	ei->i_flags |= EXT4_EXTENTS_FL;
+	ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS);
 	memcpy(ei->i_data, tmp_ei->i_data, sizeof(ei->i_data));
 
 	/*
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 52abfa12762..5f1ed9fc913 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -148,17 +148,17 @@ mext_next_extent(struct inode *inode, struct ext4_ext_path *path,
  */
 static int
 mext_check_null_inode(struct inode *inode1, struct inode *inode2,
-		const char *function)
+		      const char *function, unsigned int line)
 {
 	int ret = 0;
 
 	if (inode1 == NULL) {
-		__ext4_error(inode2->i_sb, function,
+		__ext4_error(inode2->i_sb, function, line,
 			"Both inodes should not be NULL: "
 			"inode1 NULL inode2 %lu", inode2->i_ino);
 		ret = -EIO;
 	} else if (inode2 == NULL) {
-		__ext4_error(inode1->i_sb, function,
+		__ext4_error(inode1->i_sb, function, line,
 			"Both inodes should not be NULL: "
 			"inode1 %lu inode2 NULL", inode1->i_ino);
 		ret = -EIO;
@@ -1084,7 +1084,7 @@ mext_inode_double_lock(struct inode *inode1, struct inode *inode2)
 
 	BUG_ON(inode1 == NULL && inode2 == NULL);
 
-	ret = mext_check_null_inode(inode1, inode2, __func__);
+	ret = mext_check_null_inode(inode1, inode2, __func__, __LINE__);
 	if (ret < 0)
 		goto out;
 
@@ -1121,7 +1121,7 @@ mext_inode_double_unlock(struct inode *inode1, struct inode *inode2)
 
 	BUG_ON(inode1 == NULL && inode2 == NULL);
 
-	ret = mext_check_null_inode(inode1, inode2, __func__);
+	ret = mext_check_null_inode(inode1, inode2, __func__, __LINE__);
 	if (ret < 0)
 		goto out;
 
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index a43e6617b35..314c0d3b3fa 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -179,30 +179,6 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir,
 static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry,
 			     struct inode *inode);
 
-unsigned int ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize)
-{
-	unsigned len = le16_to_cpu(dlen);
-
-	if (len == EXT4_MAX_REC_LEN || len == 0)
-		return blocksize;
-	return (len & 65532) | ((len & 3) << 16);
-}
-
-__le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize)
-{
-	if ((len > blocksize) || (blocksize > (1 << 18)) || (len & 3))
-		BUG();
-	if (len < 65536)
-		return cpu_to_le16(len);
-	if (len == blocksize) {
-		if (blocksize == 65536)
-			return cpu_to_le16(EXT4_MAX_REC_LEN);
-		else
-			return cpu_to_le16(0);
-	}
-	return cpu_to_le16((len & 65532) | ((len >> 16) & 3));
-}
-
 /*
  * p is at least 6 bytes before the end of page
  */
@@ -605,7 +581,7 @@ static int htree_dirblock_to_tree(struct file *dir_file,
 					   dir->i_sb->s_blocksize -
 					   EXT4_DIR_REC_LEN(0));
 	for (; de < top; de = ext4_next_entry(de, dir->i_sb->s_blocksize)) {
-		if (!ext4_check_dir_entry("htree_dirblock_to_tree", dir, de, bh,
+		if (!ext4_check_dir_entry(dir, de, bh,
 					(block<<EXT4_BLOCK_SIZE_BITS(dir->i_sb))
 						+((char *)de - bh->b_data))) {
 			/* On error, skip the f_pos to the next block. */
@@ -844,8 +820,7 @@ static inline int search_dirblock(struct buffer_head *bh,
 		if ((char *) de + namelen <= dlimit &&
 		    ext4_match (namelen, name, de)) {
 			/* found a match - just to be sure, do a full check */
-			if (!ext4_check_dir_entry("ext4_find_entry",
-						  dir, de, bh, offset))
+			if (!ext4_check_dir_entry(dir, de, bh, offset))
 				return -1;
 			*res_dir = de;
 			return 1;
@@ -1019,7 +994,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q
 			int off = (block << EXT4_BLOCK_SIZE_BITS(sb))
 				  + ((char *) de - bh->b_data);
 
-			if (!ext4_check_dir_entry(__func__, dir, de, bh, off)) {
+			if (!ext4_check_dir_entry(dir, de, bh, off)) {
 				brelse(bh);
 				*err = ERR_BAD_DX_DIR;
 				goto errout;
@@ -1088,7 +1063,6 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru
 struct dentry *ext4_get_parent(struct dentry *child)
 {
 	__u32 ino;
-	struct inode *inode;
 	static const struct qstr dotdot = {
 		.name = "..",
 		.len = 2,
@@ -1097,7 +1071,6 @@ struct dentry *ext4_get_parent(struct dentry *child)
 	struct buffer_head *bh;
 
 	bh = ext4_find_entry(child->d_inode, &dotdot, &de);
-	inode = NULL;
 	if (!bh)
 		return ERR_PTR(-ENOENT);
 	ino = le32_to_cpu(de->inode);
@@ -1305,8 +1278,7 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry,
 		de = (struct ext4_dir_entry_2 *)bh->b_data;
 		top = bh->b_data + blocksize - reclen;
 		while ((char *) de <= top) {
-			if (!ext4_check_dir_entry("ext4_add_entry", dir, de,
-						  bh, offset))
+			if (!ext4_check_dir_entry(dir, de, bh, offset))
 				return -EIO;
 			if (ext4_match(namelen, name, de))
 				return -EEXIST;
@@ -1673,7 +1645,7 @@ static int ext4_delete_entry(handle_t *handle,
 	pde = NULL;
 	de = (struct ext4_dir_entry_2 *) bh->b_data;
 	while (i < bh->b_size) {
-		if (!ext4_check_dir_entry("ext4_delete_entry", dir, de, bh, i))
+		if (!ext4_check_dir_entry(dir, de, bh, i))
 			return -EIO;
 		if (de == de_del)  {
 			BUFFER_TRACE(bh, "get_write_access");
@@ -1956,7 +1928,7 @@ static int empty_dir(struct inode *inode)
 			}
 			de = (struct ext4_dir_entry_2 *) bh->b_data;
 		}
-		if (!ext4_check_dir_entry("empty_dir", inode, de, bh, offset)) {
+		if (!ext4_check_dir_entry(inode, de, bh, offset)) {
 			de = (struct ext4_dir_entry_2 *)(bh->b_data +
 							 sb->s_blocksize);
 			offset = (offset | (sb->s_blocksize - 1)) + 1;
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 6df797eb9ae..ca5c8aa00a2 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -921,8 +921,7 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 			   &sbi->s_flex_groups[flex_group].free_inodes);
 	}
 
-	ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
-	sb->s_dirt = 1;
+	ext4_handle_dirty_super(handle, sb);
 
 exit_journal:
 	mutex_unlock(&sbi->s_resize_lock);
@@ -953,7 +952,6 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 		      ext4_fsblk_t n_blocks_count)
 {
 	ext4_fsblk_t o_blocks_count;
-	ext4_group_t o_groups_count;
 	ext4_grpblk_t last;
 	ext4_grpblk_t add;
 	struct buffer_head *bh;
@@ -965,7 +963,6 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 	 * yet: we're going to revalidate es->s_blocks_count after
 	 * taking the s_resize_lock below. */
 	o_blocks_count = ext4_blocks_count(es);
-	o_groups_count = EXT4_SB(sb)->s_groups_count;
 
 	if (test_opt(sb, DEBUG))
 		printk(KERN_DEBUG "EXT4-fs: extending last group from %llu uto %llu blocks\n",
@@ -1045,13 +1042,12 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 		goto exit_put;
 	}
 	ext4_blocks_count_set(es, o_blocks_count + add);
-	ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
-	sb->s_dirt = 1;
 	mutex_unlock(&EXT4_SB(sb)->s_resize_lock);
 	ext4_debug("freeing blocks %llu through %llu\n", o_blocks_count,
 		   o_blocks_count + add);
 	/* We add the blocks to the bitmap and set the group need init bit */
 	ext4_add_groupblocks(handle, sb, o_blocks_count, add);
+	ext4_handle_dirty_super(handle, sb);
 	ext4_debug("freed blocks %llu through %llu\n", o_blocks_count,
 		   o_blocks_count + add);
 	if ((err = ext4_journal_stop(handle)))
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index e72d3235b2f..8d65575f8c8 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -241,14 +241,14 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
 	if (sb->s_flags & MS_RDONLY)
 		return ERR_PTR(-EROFS);
 
-	vfs_check_frozen(sb, SB_FREEZE_WRITE);
+	vfs_check_frozen(sb, SB_FREEZE_TRANS);
 	/* Special case here: if the journal has aborted behind our
 	 * backs (eg. EIO in the commit thread), then we still need to
 	 * take the FS itself readonly cleanly. */
 	journal = EXT4_SB(sb)->s_journal;
 	if (journal) {
 		if (is_journal_aborted(journal)) {
-			ext4_abort(sb, __func__, "Detected aborted journal");
+			ext4_abort(sb, "Detected aborted journal");
 			return ERR_PTR(-EROFS);
 		}
 		return jbd2_journal_start(journal, nblocks);
@@ -262,7 +262,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
  * that sync() will call the filesystem's write_super callback if
  * appropriate.
  */
-int __ext4_journal_stop(const char *where, handle_t *handle)
+int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle)
 {
 	struct super_block *sb;
 	int err;
@@ -279,12 +279,13 @@ int __ext4_journal_stop(const char *where, handle_t *handle)
 	if (!err)
 		err = rc;
 	if (err)
-		__ext4_std_error(sb, where, err);
+		__ext4_std_error(sb, where, line, err);
 	return err;
 }
 
-void ext4_journal_abort_handle(const char *caller, const char *err_fn,
-		struct buffer_head *bh, handle_t *handle, int err)
+void ext4_journal_abort_handle(const char *caller, unsigned int line,
+			       const char *err_fn, struct buffer_head *bh,
+			       handle_t *handle, int err)
 {
 	char nbuf[16];
 	const char *errstr = ext4_decode_error(NULL, err, nbuf);
@@ -300,12 +301,47 @@ void ext4_journal_abort_handle(const char *caller, const char *err_fn,
 	if (is_handle_aborted(handle))
 		return;
 
-	printk(KERN_ERR "%s: aborting transaction: %s in %s\n",
-	       caller, errstr, err_fn);
+	printk(KERN_ERR "%s:%d: aborting transaction: %s in %s\n",
+	       caller, line, errstr, err_fn);
 
 	jbd2_journal_abort_handle(handle);
 }
 
+static void __save_error_info(struct super_block *sb, const char *func,
+			    unsigned int line)
+{
+	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
+
+	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
+	es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
+	es->s_last_error_time = cpu_to_le32(get_seconds());
+	strncpy(es->s_last_error_func, func, sizeof(es->s_last_error_func));
+	es->s_last_error_line = cpu_to_le32(line);
+	if (!es->s_first_error_time) {
+		es->s_first_error_time = es->s_last_error_time;
+		strncpy(es->s_first_error_func, func,
+			sizeof(es->s_first_error_func));
+		es->s_first_error_line = cpu_to_le32(line);
+		es->s_first_error_ino = es->s_last_error_ino;
+		es->s_first_error_block = es->s_last_error_block;
+	}
+	/*
+	 * Start the daily error reporting function if it hasn't been
+	 * started already
+	 */
+	if (!es->s_error_count)
+		mod_timer(&EXT4_SB(sb)->s_err_report, jiffies + 24*60*60*HZ);
+	es->s_error_count = cpu_to_le32(le32_to_cpu(es->s_error_count) + 1);
+}
+
+static void save_error_info(struct super_block *sb, const char *func,
+			    unsigned int line)
+{
+	__save_error_info(sb, func, line);
+	ext4_commit_super(sb, 1);
+}
+
+
 /* Deal with the reporting of failure conditions on a filesystem such as
  * inconsistencies detected or read IO failures.
  *
@@ -323,11 +359,6 @@ void ext4_journal_abort_handle(const char *caller, const char *err_fn,
 
 static void ext4_handle_error(struct super_block *sb)
 {
-	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
-
-	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
-	es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
-
 	if (sb->s_flags & MS_RDONLY)
 		return;
 
@@ -342,19 +373,19 @@ static void ext4_handle_error(struct super_block *sb)
 		ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only");
 		sb->s_flags |= MS_RDONLY;
 	}
-	ext4_commit_super(sb, 1);
 	if (test_opt(sb, ERRORS_PANIC))
 		panic("EXT4-fs (device %s): panic forced after error\n",
 			sb->s_id);
 }
 
 void __ext4_error(struct super_block *sb, const char *function,
-		const char *fmt, ...)
+		  unsigned int line, const char *fmt, ...)
 {
 	va_list args;
 
 	va_start(args, fmt);
-	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
+	printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: comm %s: ",
+	       sb->s_id, function, line, current->comm);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
@@ -362,14 +393,22 @@ void __ext4_error(struct super_block *sb, const char *function,
 	ext4_handle_error(sb);
 }
 
-void ext4_error_inode(const char *function, struct inode *inode,
+void ext4_error_inode(struct inode *inode, const char *function,
+		      unsigned int line, ext4_fsblk_t block,
 		      const char *fmt, ...)
 {
 	va_list args;
+	struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
 
+	es->s_last_error_ino = cpu_to_le32(inode->i_ino);
+	es->s_last_error_block = cpu_to_le64(block);
+	save_error_info(inode->i_sb, function, line);
 	va_start(args, fmt);
-	printk(KERN_CRIT "EXT4-fs error (device %s): %s: inode #%lu: (comm %s) ",
-	       inode->i_sb->s_id, function, inode->i_ino, current->comm);
+	printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: inode #%lu: ",
+	       inode->i_sb->s_id, function, line, inode->i_ino);
+	if (block)
+		printk("block %llu: ", block);
+	printk("comm %s: ", current->comm);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
@@ -377,20 +416,26 @@ void ext4_error_inode(const char *function, struct inode *inode,
 	ext4_handle_error(inode->i_sb);
 }
 
-void ext4_error_file(const char *function, struct file *file,
-		     const char *fmt, ...)
+void ext4_error_file(struct file *file, const char *function,
+		     unsigned int line, const char *fmt, ...)
 {
 	va_list args;
+	struct ext4_super_block *es;
 	struct inode *inode = file->f_dentry->d_inode;
 	char pathname[80], *path;
 
+	es = EXT4_SB(inode->i_sb)->s_es;
+	es->s_last_error_ino = cpu_to_le32(inode->i_ino);
+	save_error_info(inode->i_sb, function, line);
 	va_start(args, fmt);
 	path = d_path(&(file->f_path), pathname, sizeof(pathname));
 	if (!path)
 		path = "(unknown)";
 	printk(KERN_CRIT
-	       "EXT4-fs error (device %s): %s: inode #%lu (comm %s path %s): ",
-	       inode->i_sb->s_id, function, inode->i_ino, current->comm, path);
+	       "EXT4-fs error (device %s): %s:%d: inode #%lu "
+	       "(comm %s path %s): ",
+	       inode->i_sb->s_id, function, line, inode->i_ino,
+	       current->comm, path);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
@@ -435,7 +480,8 @@ static const char *ext4_decode_error(struct super_block *sb, int errno,
 /* __ext4_std_error decodes expected errors from journaling functions
  * automatically and invokes the appropriate error response.  */
 
-void __ext4_std_error(struct super_block *sb, const char *function, int errno)
+void __ext4_std_error(struct super_block *sb, const char *function,
+		      unsigned int line, int errno)
 {
 	char nbuf[16];
 	const char *errstr;
@@ -448,8 +494,9 @@ void __ext4_std_error(struct super_block *sb, const char *function, int errno)
 		return;
 
 	errstr = ext4_decode_error(sb, errno, nbuf);
-	printk(KERN_CRIT "EXT4-fs error (device %s) in %s: %s\n",
-	       sb->s_id, function, errstr);
+	printk(KERN_CRIT "EXT4-fs error (device %s) in %s:%d: %s\n",
+	       sb->s_id, function, line, errstr);
+	save_error_info(sb, function, line);
 
 	ext4_handle_error(sb);
 }
@@ -464,29 +511,29 @@ void __ext4_std_error(struct super_block *sb, const char *function, int errno)
  * case we take the easy way out and panic immediately.
  */
 
-void ext4_abort(struct super_block *sb, const char *function,
-		const char *fmt, ...)
+void __ext4_abort(struct super_block *sb, const char *function,
+		unsigned int line, const char *fmt, ...)
 {
 	va_list args;
 
+	save_error_info(sb, function, line);
 	va_start(args, fmt);
-	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
+	printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: ", sb->s_id,
+	       function, line);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
 
+	if ((sb->s_flags & MS_RDONLY) == 0) {
+		ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only");
+		sb->s_flags |= MS_RDONLY;
+		EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED;
+		if (EXT4_SB(sb)->s_journal)
+			jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO);
+		save_error_info(sb, function, line);
+	}
 	if (test_opt(sb, ERRORS_PANIC))
 		panic("EXT4-fs panic from previous error\n");
-
-	if (sb->s_flags & MS_RDONLY)
-		return;
-
-	ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only");
-	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
-	sb->s_flags |= MS_RDONLY;
-	EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED;
-	if (EXT4_SB(sb)->s_journal)
-		jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO);
 }
 
 void ext4_msg (struct super_block * sb, const char *prefix,
@@ -502,38 +549,47 @@ void ext4_msg (struct super_block * sb, const char *prefix,
 }
 
 void __ext4_warning(struct super_block *sb, const char *function,
-		  const char *fmt, ...)
+		    unsigned int line, const char *fmt, ...)
 {
 	va_list args;
 
 	va_start(args, fmt);
-	printk(KERN_WARNING "EXT4-fs warning (device %s): %s: ",
-	       sb->s_id, function);
+	printk(KERN_WARNING "EXT4-fs warning (device %s): %s:%d: ",
+	       sb->s_id, function, line);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
 }
 
-void ext4_grp_locked_error(struct super_block *sb, ext4_group_t grp,
-			   const char *function, const char *fmt, ...)
+void __ext4_grp_locked_error(const char *function, unsigned int line,
+			     struct super_block *sb, ext4_group_t grp,
+			     unsigned long ino, ext4_fsblk_t block,
+			     const char *fmt, ...)
 __releases(bitlock)
 __acquires(bitlock)
 {
 	va_list args;
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 
+	es->s_last_error_ino = cpu_to_le32(ino);
+	es->s_last_error_block = cpu_to_le64(block);
+	__save_error_info(sb, function, line);
 	va_start(args, fmt);
-	printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function);
+	printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: group %u",
+	       sb->s_id, function, line, grp);
+	if (ino)
+		printk("inode %lu: ", ino);
+	if (block)
+		printk("block %llu:", (unsigned long long) block);
 	vprintk(fmt, args);
 	printk("\n");
 	va_end(args);
 
 	if (test_opt(sb, ERRORS_CONT)) {
-		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
-		es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
 		ext4_commit_super(sb, 0);
 		return;
 	}
+
 	ext4_unlock_group(sb, grp);
 	ext4_handle_error(sb);
 	/*
@@ -660,8 +716,7 @@ static void ext4_put_super(struct super_block *sb)
 		err = jbd2_journal_destroy(sbi->s_journal);
 		sbi->s_journal = NULL;
 		if (err < 0)
-			ext4_abort(sb, __func__,
-				   "Couldn't clean up the journal");
+			ext4_abort(sb, "Couldn't clean up the journal");
 	}
 
 	ext4_release_system_zone(sb);
@@ -946,14 +1001,12 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 		seq_puts(seq, ",journal_async_commit");
 	else if (test_opt(sb, JOURNAL_CHECKSUM))
 		seq_puts(seq, ",journal_checksum");
-	if (test_opt(sb, NOBH))
-		seq_puts(seq, ",nobh");
 	if (test_opt(sb, I_VERSION))
 		seq_puts(seq, ",i_version");
-	if (!test_opt(sb, DELALLOC))
+	if (!test_opt(sb, DELALLOC) &&
+	    !(def_mount_opts & EXT4_DEFM_NODELALLOC))
 		seq_puts(seq, ",nodelalloc");
 
-
 	if (sbi->s_stripe)
 		seq_printf(seq, ",stripe=%lu", sbi->s_stripe);
 	/*
@@ -977,7 +1030,7 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (test_opt(sb, NO_AUTO_DA_ALLOC))
 		seq_puts(seq, ",noauto_da_alloc");
 
-	if (test_opt(sb, DISCARD))
+	if (test_opt(sb, DISCARD) && !(def_mount_opts & EXT4_DEFM_DISCARD))
 		seq_puts(seq, ",discard");
 
 	if (test_opt(sb, NOLOAD))
@@ -986,6 +1039,10 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (test_opt(sb, DIOREAD_NOLOCK))
 		seq_puts(seq, ",dioread_nolock");
 
+	if (test_opt(sb, BLOCK_VALIDITY) &&
+	    !(def_mount_opts & EXT4_DEFM_BLOCK_VALIDITY))
+		seq_puts(seq, ",block_validity");
+
 	ext4_show_quota_options(seq, sb);
 
 	return 0;
@@ -1065,6 +1122,7 @@ static int ext4_mark_dquot_dirty(struct dquot *dquot);
 static int ext4_write_info(struct super_block *sb, int type);
 static int ext4_quota_on(struct super_block *sb, int type, int format_id,
 				char *path);
+static int ext4_quota_off(struct super_block *sb, int type);
 static int ext4_quota_on_mount(struct super_block *sb, int type);
 static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data,
 			       size_t len, loff_t off);
@@ -1086,7 +1144,7 @@ static const struct dquot_operations ext4_quota_operations = {
 
 static const struct quotactl_ops ext4_qctl_operations = {
 	.quota_on	= ext4_quota_on,
-	.quota_off	= dquot_quota_off,
+	.quota_off	= ext4_quota_off,
 	.quota_sync	= dquot_quota_sync,
 	.get_info	= dquot_get_dqinfo,
 	.set_info	= dquot_set_dqinfo,
@@ -1624,10 +1682,12 @@ set_qf_format:
 			*n_blocks_count = option;
 			break;
 		case Opt_nobh:
-			set_opt(sbi->s_mount_opt, NOBH);
+			ext4_msg(sb, KERN_WARNING,
+				 "Ignoring deprecated nobh option");
 			break;
 		case Opt_bh:
-			clear_opt(sbi->s_mount_opt, NOBH);
+			ext4_msg(sb, KERN_WARNING,
+				 "Ignoring deprecated bh option");
 			break;
 		case Opt_i_version:
 			set_opt(sbi->s_mount_opt, I_VERSION);
@@ -2249,6 +2309,8 @@ static ssize_t session_write_kbytes_show(struct ext4_attr *a,
 {
 	struct super_block *sb = sbi->s_buddy_cache->i_sb;
 
+	if (!sb->s_bdev->bd_part)
+		return snprintf(buf, PAGE_SIZE, "0\n");
 	return snprintf(buf, PAGE_SIZE, "%lu\n",
 			(part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
 			 sbi->s_sectors_written_start) >> 1);
@@ -2259,6 +2321,8 @@ static ssize_t lifetime_write_kbytes_show(struct ext4_attr *a,
 {
 	struct super_block *sb = sbi->s_buddy_cache->i_sb;
 
+	if (!sb->s_bdev->bd_part)
+		return snprintf(buf, PAGE_SIZE, "0\n");
 	return snprintf(buf, PAGE_SIZE, "%llu\n",
 			(unsigned long long)(sbi->s_kbytes_written +
 			((part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
@@ -2431,6 +2495,53 @@ static int ext4_feature_set_ok(struct super_block *sb, int readonly)
 	return 1;
 }
 
+/*
+ * This function is called once a day if we have errors logged
+ * on the file system
+ */
+static void print_daily_error_info(unsigned long arg)
+{
+	struct super_block *sb = (struct super_block *) arg;
+	struct ext4_sb_info *sbi;
+	struct ext4_super_block *es;
+
+	sbi = EXT4_SB(sb);
+	es = sbi->s_es;
+
+	if (es->s_error_count)
+		ext4_msg(sb, KERN_NOTICE, "error count: %u",
+			 le32_to_cpu(es->s_error_count));
+	if (es->s_first_error_time) {
+		printk(KERN_NOTICE "EXT4-fs (%s): initial error at %u: %.*s:%d",
+		       sb->s_id, le32_to_cpu(es->s_first_error_time),
+		       (int) sizeof(es->s_first_error_func),
+		       es->s_first_error_func,
+		       le32_to_cpu(es->s_first_error_line));
+		if (es->s_first_error_ino)
+			printk(": inode %u",
+			       le32_to_cpu(es->s_first_error_ino));
+		if (es->s_first_error_block)
+			printk(": block %llu", (unsigned long long)
+			       le64_to_cpu(es->s_first_error_block));
+		printk("\n");
+	}
+	if (es->s_last_error_time) {
+		printk(KERN_NOTICE "EXT4-fs (%s): last error at %u: %.*s:%d",
+		       sb->s_id, le32_to_cpu(es->s_last_error_time),
+		       (int) sizeof(es->s_last_error_func),
+		       es->s_last_error_func,
+		       le32_to_cpu(es->s_last_error_line));
+		if (es->s_last_error_ino)
+			printk(": inode %u",
+			       le32_to_cpu(es->s_last_error_ino));
+		if (es->s_last_error_block)
+			printk(": block %llu", (unsigned long long)
+			       le64_to_cpu(es->s_last_error_block));
+		printk("\n");
+	}
+	mod_timer(&sbi->s_err_report, jiffies + 24*60*60*HZ);  /* Once a day */
+}
+
 static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 				__releases(kernel_lock)
 				__acquires(kernel_lock)
@@ -2448,7 +2559,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	struct inode *root;
 	char *cp;
 	const char *descr;
-	int ret = -EINVAL;
+	int ret = -ENOMEM;
 	int blocksize;
 	unsigned int db_count;
 	unsigned int i;
@@ -2459,13 +2570,13 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 
 	sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
 	if (!sbi)
-		return -ENOMEM;
+		goto out_free_orig;
 
 	sbi->s_blockgroup_lock =
 		kzalloc(sizeof(struct blockgroup_lock), GFP_KERNEL);
 	if (!sbi->s_blockgroup_lock) {
 		kfree(sbi);
-		return -ENOMEM;
+		goto out_free_orig;
 	}
 	sb->s_fs_info = sbi;
 	sbi->s_mount_opt = 0;
@@ -2473,8 +2584,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	sbi->s_resgid = EXT4_DEF_RESGID;
 	sbi->s_inode_readahead_blks = EXT4_DEF_INODE_READAHEAD_BLKS;
 	sbi->s_sb_block = sb_block;
-	sbi->s_sectors_written_start = part_stat_read(sb->s_bdev->bd_part,
-						      sectors[1]);
+	if (sb->s_bdev->bd_part)
+		sbi->s_sectors_written_start =
+			part_stat_read(sb->s_bdev->bd_part, sectors[1]);
 
 	unlock_kernel();
 
@@ -2482,6 +2594,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	for (cp = sb->s_id; (cp = strchr(cp, '/'));)
 		*cp = '!';
 
+	ret = -EINVAL;
 	blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE);
 	if (!blocksize) {
 		ext4_msg(sb, KERN_ERR, "unable to set blocksize");
@@ -2546,6 +2659,10 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		set_opt(sbi->s_mount_opt, ERRORS_CONT);
 	else
 		set_opt(sbi->s_mount_opt, ERRORS_RO);
+	if (def_mount_opts & EXT4_DEFM_BLOCK_VALIDITY)
+		set_opt(sbi->s_mount_opt, BLOCK_VALIDITY);
+	if (def_mount_opts & EXT4_DEFM_DISCARD)
+		set_opt(sbi->s_mount_opt, DISCARD);
 
 	sbi->s_resuid = le16_to_cpu(es->s_def_resuid);
 	sbi->s_resgid = le16_to_cpu(es->s_def_resgid);
@@ -2553,15 +2670,23 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	sbi->s_min_batch_time = EXT4_DEF_MIN_BATCH_TIME;
 	sbi->s_max_batch_time = EXT4_DEF_MAX_BATCH_TIME;
 
-	set_opt(sbi->s_mount_opt, BARRIER);
+	if ((def_mount_opts & EXT4_DEFM_NOBARRIER) == 0)
+		set_opt(sbi->s_mount_opt, BARRIER);
 
 	/*
 	 * enable delayed allocation by default
 	 * Use -o nodelalloc to turn it off
 	 */
-	if (!IS_EXT3_SB(sb))
+	if (!IS_EXT3_SB(sb) &&
+	    ((def_mount_opts & EXT4_DEFM_NODELALLOC) == 0))
 		set_opt(sbi->s_mount_opt, DELALLOC);
 
+	if (!parse_options((char *) sbi->s_es->s_mount_opts, sb,
+			   &journal_devnum, &journal_ioprio, NULL, 0)) {
+		ext4_msg(sb, KERN_WARNING,
+			 "failed to parse options in superblock: %s",
+			 sbi->s_es->s_mount_opts);
+	}
 	if (!parse_options((char *) data, sb, &journal_devnum,
 			   &journal_ioprio, NULL, 0))
 		goto failed_mount;
@@ -2912,18 +3037,7 @@ no_journal:
 		ext4_msg(sb, KERN_ERR, "insufficient memory");
 		goto failed_mount_wq;
 	}
-	if (test_opt(sb, NOBH)) {
-		if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) {
-			ext4_msg(sb, KERN_WARNING, "Ignoring nobh option - "
-				"its supported only with writeback mode");
-			clear_opt(sbi->s_mount_opt, NOBH);
-		}
-		if (test_opt(sb, DIOREAD_NOLOCK)) {
-			ext4_msg(sb, KERN_WARNING, "dioread_nolock option is "
-				"not supported with nobh mode");
-			goto failed_mount_wq;
-		}
-	}
+
 	EXT4_SB(sb)->dio_unwritten_wq = create_workqueue("ext4-dio-unwritten");
 	if (!EXT4_SB(sb)->dio_unwritten_wq) {
 		printk(KERN_ERR "EXT4-fs: failed to create DIO workqueue\n");
@@ -3043,7 +3157,14 @@ no_journal:
 		descr = "out journal";
 
 	ext4_msg(sb, KERN_INFO, "mounted filesystem with%s. "
-		"Opts: %s", descr, orig_data);
+		 "Opts: %s%s%s", descr, sbi->s_es->s_mount_opts,
+		 *sbi->s_es->s_mount_opts ? "; " : "", orig_data);
+
+	init_timer(&sbi->s_err_report);
+	sbi->s_err_report.function = print_daily_error_info;
+	sbi->s_err_report.data = (unsigned long) sb;
+	if (es->s_error_count)
+		mod_timer(&sbi->s_err_report, jiffies + 300*HZ); /* 5 minutes */
 
 	lock_kernel();
 	kfree(orig_data);
@@ -3093,6 +3214,7 @@ out_fail:
 	kfree(sbi->s_blockgroup_lock);
 	kfree(sbi);
 	lock_kernel();
+out_free_orig:
 	kfree(orig_data);
 	return ret;
 }
@@ -3110,7 +3232,7 @@ static void ext4_init_journal_params(struct super_block *sb, journal_t *journal)
 	journal->j_min_batch_time = sbi->s_min_batch_time;
 	journal->j_max_batch_time = sbi->s_max_batch_time;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	if (test_opt(sb, BARRIER))
 		journal->j_flags |= JBD2_BARRIER;
 	else
@@ -3119,7 +3241,7 @@ static void ext4_init_journal_params(struct super_block *sb, journal_t *journal)
 		journal->j_flags |= JBD2_ABORT_ON_SYNCDATA_ERR;
 	else
 		journal->j_flags &= ~JBD2_ABORT_ON_SYNCDATA_ERR;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 static journal_t *ext4_get_journal(struct super_block *sb,
@@ -3327,8 +3449,17 @@ static int ext4_load_journal(struct super_block *sb,
 
 	if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER))
 		err = jbd2_journal_wipe(journal, !really_read_only);
-	if (!err)
+	if (!err) {
+		char *save = kmalloc(EXT4_S_ERR_LEN, GFP_KERNEL);
+		if (save)
+			memcpy(save, ((char *) es) +
+			       EXT4_S_ERR_START, EXT4_S_ERR_LEN);
 		err = jbd2_journal_load(journal);
+		if (save)
+			memcpy(((char *) es) + EXT4_S_ERR_START,
+			       save, EXT4_S_ERR_LEN);
+		kfree(save);
+	}
 
 	if (err) {
 		ext4_msg(sb, KERN_ERR, "error loading journal");
@@ -3384,10 +3515,14 @@ static int ext4_commit_super(struct super_block *sb, int sync)
 	 */
 	if (!(sb->s_flags & MS_RDONLY))
 		es->s_wtime = cpu_to_le32(get_seconds());
-	es->s_kbytes_written =
-		cpu_to_le64(EXT4_SB(sb)->s_kbytes_written +
+	if (sb->s_bdev->bd_part)
+		es->s_kbytes_written =
+			cpu_to_le64(EXT4_SB(sb)->s_kbytes_written +
 			    ((part_stat_read(sb->s_bdev->bd_part, sectors[1]) -
 			      EXT4_SB(sb)->s_sectors_written_start) >> 1));
+	else
+		es->s_kbytes_written =
+			cpu_to_le64(EXT4_SB(sb)->s_kbytes_written);
 	ext4_free_blocks_count_set(es, percpu_counter_sum_positive(
 					&EXT4_SB(sb)->s_freeblocks_counter));
 	es->s_free_inodes_count = cpu_to_le32(percpu_counter_sum_positive(
@@ -3491,7 +3626,7 @@ int ext4_force_commit(struct super_block *sb)
 
 	journal = EXT4_SB(sb)->s_journal;
 	if (journal) {
-		vfs_check_frozen(sb, SB_FREEZE_WRITE);
+		vfs_check_frozen(sb, SB_FREEZE_TRANS);
 		ret = ext4_journal_force_commit(journal);
 	}
 
@@ -3616,7 +3751,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 	}
 
 	if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED)
-		ext4_abort(sb, __func__, "Abort forced by user");
+		ext4_abort(sb, "Abort forced by user");
 
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
 		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
@@ -3981,6 +4116,18 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id,
 	return err;
 }
 
+static int ext4_quota_off(struct super_block *sb, int type)
+{
+	/* Force all delayed allocation blocks to be allocated */
+	if (test_opt(sb, DELALLOC)) {
+		down_read(&sb->s_umount);
+		sync_filesystem(sb);
+		up_read(&sb->s_umount);
+	}
+
+	return dquot_quota_off(sb, type);
+}
+
 /* Read data from quotafile - avoid pagecache and such because we cannot afford
  * acquiring the locks... As quota files are never truncated and quota code
  * itself serializes the operations (and noone else should touch the files)
@@ -4030,7 +4177,6 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
 	ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
 	int err = 0;
 	int offset = off & (sb->s_blocksize - 1);
-	int journal_quota = EXT4_SB(sb)->s_qf_names[type] != NULL;
 	struct buffer_head *bh;
 	handle_t *handle = journal_current_handle();
 
@@ -4055,24 +4201,16 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
 	bh = ext4_bread(handle, inode, blk, 1, &err);
 	if (!bh)
 		goto out;
-	if (journal_quota) {
-		err = ext4_journal_get_write_access(handle, bh);
-		if (err) {
-			brelse(bh);
-			goto out;
-		}
+	err = ext4_journal_get_write_access(handle, bh);
+	if (err) {
+		brelse(bh);
+		goto out;
 	}
 	lock_buffer(bh);
 	memcpy(bh->b_data+offset, data, len);
 	flush_dcache_page(bh->b_page);
 	unlock_buffer(bh);
-	if (journal_quota)
-		err = ext4_handle_dirty_metadata(handle, NULL, bh);
-	else {
-		/* Always do at least ordered writes for quotas */
-		err = ext4_jbd2_file_inode(handle, inode);
-		mark_buffer_dirty(bh);
-	}
+	err = ext4_handle_dirty_metadata(handle, NULL, bh);
 	brelse(bh);
 out:
 	if (err) {
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 04338009793..a6f31424957 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -458,8 +458,7 @@ static void ext4_xattr_update_super_block(handle_t *handle,
 
 	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
 		EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR);
-		sb->s_dirt = 1;
-		ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
+		ext4_handle_dirty_super(handle, sb);
 	}
 }
 
diff --git a/fs/file.c b/fs/file.c
index 34bb7f71d99..cccaead962c 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -178,7 +178,6 @@ static struct fdtable * alloc_fdtable(unsigned int nr)
 	fdt->open_fds = (fd_set *)data;
 	data += nr / BITS_PER_BYTE;
 	fdt->close_on_exec = (fd_set *)data;
-	INIT_RCU_HEAD(&fdt->rcu);
 	fdt->next = NULL;
 
 	return fdt;
@@ -312,7 +311,6 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
 	new_fdt->close_on_exec = (fd_set *)&newf->close_on_exec_init;
 	new_fdt->open_fds = (fd_set *)&newf->open_fds_init;
 	new_fdt->fd = &newf->fd_array[0];
-	INIT_RCU_HEAD(&new_fdt->rcu);
 	new_fdt->next = NULL;
 
 	spin_lock(&oldf->file_lock);
@@ -430,7 +428,6 @@ struct files_struct init_files = {
 		.fd		= &init_files.fd_array[0],
 		.close_on_exec	= (fd_set *)&init_files.close_on_exec_init,
 		.open_fds	= (fd_set *)&init_files.open_fds_init,
-		.rcu		= RCU_HEAD_INIT,
 	},
 	.file_lock	= __SPIN_LOCK_UNLOCKED(init_task.file_lock),
 };
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index d5be1693ac9..30ac305e829 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -530,7 +530,8 @@ void writeback_inodes_wb(struct bdi_writeback *wb,
 {
 	int ret = 0;
 
-	wbc->wb_start = jiffies; /* livelock avoidance */
+	if (!wbc->wb_start)
+		wbc->wb_start = jiffies; /* livelock avoidance */
 	spin_lock(&inode_lock);
 	if (!wbc->for_kupdate || list_empty(&wb->b_io))
 		queue_io(wb, wbc->older_than_this);
@@ -559,7 +560,6 @@ static void __writeback_inodes_sb(struct super_block *sb,
 {
 	WARN_ON(!rwsem_is_locked(&sb->s_umount));
 
-	wbc->wb_start = jiffies; /* livelock avoidance */
 	spin_lock(&inode_lock);
 	if (!wbc->for_kupdate || list_empty(&wb->b_io))
 		queue_io(wb, wbc->older_than_this);
@@ -625,6 +625,7 @@ static long wb_writeback(struct bdi_writeback *wb,
 		wbc.range_end = LLONG_MAX;
 	}
 
+	wbc.wb_start = jiffies; /* livelock avoidance */
 	for (;;) {
 		/*
 		 * Stop writeback when nr_pages has been consumed
diff --git a/fs/fscache/Kconfig b/fs/fscache/Kconfig
index cc94bb9563f..3f6dfa98988 100644
--- a/fs/fscache/Kconfig
+++ b/fs/fscache/Kconfig
@@ -1,7 +1,6 @@
 
 config FSCACHE
 	tristate "General filesystem local caching manager"
-	select SLOW_WORK
 	help
 	  This option enables a generic filesystem caching manager that can be
 	  used by various network and other filesystems to cache data locally.
diff --git a/fs/fscache/internal.h b/fs/fscache/internal.h
index edd7434ab6e..6a026441c5a 100644
--- a/fs/fscache/internal.h
+++ b/fs/fscache/internal.h
@@ -82,6 +82,14 @@ extern unsigned fscache_defer_lookup;
 extern unsigned fscache_defer_create;
 extern unsigned fscache_debug;
 extern struct kobject *fscache_root;
+extern struct workqueue_struct *fscache_object_wq;
+extern struct workqueue_struct *fscache_op_wq;
+DECLARE_PER_CPU(wait_queue_head_t, fscache_object_cong_wait);
+
+static inline bool fscache_object_congested(void)
+{
+	return workqueue_congested(WORK_CPU_UNBOUND, fscache_object_wq);
+}
 
 extern int fscache_wait_bit(void *);
 extern int fscache_wait_bit_interruptible(void *);
diff --git a/fs/fscache/main.c b/fs/fscache/main.c
index add6bdb53f0..f9d856773f7 100644
--- a/fs/fscache/main.c
+++ b/fs/fscache/main.c
@@ -15,6 +15,7 @@
 #include <linux/sched.h>
 #include <linux/completion.h>
 #include <linux/slab.h>
+#include <linux/seq_file.h>
 #include "internal.h"
 
 MODULE_DESCRIPTION("FS Cache Manager");
@@ -40,22 +41,105 @@ MODULE_PARM_DESC(fscache_debug,
 		 "FS-Cache debugging mask");
 
 struct kobject *fscache_root;
+struct workqueue_struct *fscache_object_wq;
+struct workqueue_struct *fscache_op_wq;
+
+DEFINE_PER_CPU(wait_queue_head_t, fscache_object_cong_wait);
+
+/* these values serve as lower bounds, will be adjusted in fscache_init() */
+static unsigned fscache_object_max_active = 4;
+static unsigned fscache_op_max_active = 2;
+
+#ifdef CONFIG_SYSCTL
+static struct ctl_table_header *fscache_sysctl_header;
+
+static int fscache_max_active_sysctl(struct ctl_table *table, int write,
+				     void __user *buffer,
+				     size_t *lenp, loff_t *ppos)
+{
+	struct workqueue_struct **wqp = table->extra1;
+	unsigned int *datap = table->data;
+	int ret;
+
+	ret = proc_dointvec(table, write, buffer, lenp, ppos);
+	if (ret == 0)
+		workqueue_set_max_active(*wqp, *datap);
+	return ret;
+}
+
+ctl_table fscache_sysctls[] = {
+	{
+		.procname	= "object_max_active",
+		.data		= &fscache_object_max_active,
+		.maxlen		= sizeof(unsigned),
+		.mode		= 0644,
+		.proc_handler	= fscache_max_active_sysctl,
+		.extra1		= &fscache_object_wq,
+	},
+	{
+		.procname	= "operation_max_active",
+		.data		= &fscache_op_max_active,
+		.maxlen		= sizeof(unsigned),
+		.mode		= 0644,
+		.proc_handler	= fscache_max_active_sysctl,
+		.extra1		= &fscache_op_wq,
+	},
+	{}
+};
+
+ctl_table fscache_sysctls_root[] = {
+	{
+		.procname	= "fscache",
+		.mode		= 0555,
+		.child		= fscache_sysctls,
+	},
+	{}
+};
+#endif
 
 /*
  * initialise the fs caching module
  */
 static int __init fscache_init(void)
 {
+	unsigned int nr_cpus = num_possible_cpus();
+	unsigned int cpu;
 	int ret;
 
-	ret = slow_work_register_user(THIS_MODULE);
-	if (ret < 0)
-		goto error_slow_work;
+	fscache_object_max_active =
+		clamp_val(nr_cpus,
+			  fscache_object_max_active, WQ_UNBOUND_MAX_ACTIVE);
+
+	ret = -ENOMEM;
+	fscache_object_wq = alloc_workqueue("fscache_object", WQ_UNBOUND,
+					    fscache_object_max_active);
+	if (!fscache_object_wq)
+		goto error_object_wq;
+
+	fscache_op_max_active =
+		clamp_val(fscache_object_max_active / 2,
+			  fscache_op_max_active, WQ_UNBOUND_MAX_ACTIVE);
+
+	ret = -ENOMEM;
+	fscache_op_wq = alloc_workqueue("fscache_operation", WQ_UNBOUND,
+					fscache_op_max_active);
+	if (!fscache_op_wq)
+		goto error_op_wq;
+
+	for_each_possible_cpu(cpu)
+		init_waitqueue_head(&per_cpu(fscache_object_cong_wait, cpu));
 
 	ret = fscache_proc_init();
 	if (ret < 0)
 		goto error_proc;
 
+#ifdef CONFIG_SYSCTL
+	ret = -ENOMEM;
+	fscache_sysctl_header = register_sysctl_table(fscache_sysctls_root);
+	if (!fscache_sysctl_header)
+		goto error_sysctl;
+#endif
+
 	fscache_cookie_jar = kmem_cache_create("fscache_cookie_jar",
 					       sizeof(struct fscache_cookie),
 					       0,
@@ -78,10 +162,16 @@ static int __init fscache_init(void)
 error_kobj:
 	kmem_cache_destroy(fscache_cookie_jar);
 error_cookie_jar:
+#ifdef CONFIG_SYSCTL
+	unregister_sysctl_table(fscache_sysctl_header);
+error_sysctl:
+#endif
 	fscache_proc_cleanup();
 error_proc:
-	slow_work_unregister_user(THIS_MODULE);
-error_slow_work:
+	destroy_workqueue(fscache_op_wq);
+error_op_wq:
+	destroy_workqueue(fscache_object_wq);
+error_object_wq:
 	return ret;
 }
 
@@ -96,8 +186,12 @@ static void __exit fscache_exit(void)
 
 	kobject_put(fscache_root);
 	kmem_cache_destroy(fscache_cookie_jar);
+#ifdef CONFIG_SYSCTL
+	unregister_sysctl_table(fscache_sysctl_header);
+#endif
 	fscache_proc_cleanup();
-	slow_work_unregister_user(THIS_MODULE);
+	destroy_workqueue(fscache_op_wq);
+	destroy_workqueue(fscache_object_wq);
 	printk(KERN_NOTICE "FS-Cache: Unloaded\n");
 }
 
diff --git a/fs/fscache/object-list.c b/fs/fscache/object-list.c
index 4a8eb31c533..ebe29c58138 100644
--- a/fs/fscache/object-list.c
+++ b/fs/fscache/object-list.c
@@ -34,8 +34,8 @@ struct fscache_objlist_data {
 #define FSCACHE_OBJLIST_CONFIG_NOREADS	0x00000200	/* show objects without active reads */
 #define FSCACHE_OBJLIST_CONFIG_EVENTS	0x00000400	/* show objects with events */
 #define FSCACHE_OBJLIST_CONFIG_NOEVENTS	0x00000800	/* show objects without no events */
-#define FSCACHE_OBJLIST_CONFIG_WORK	0x00001000	/* show objects with slow work */
-#define FSCACHE_OBJLIST_CONFIG_NOWORK	0x00002000	/* show objects without slow work */
+#define FSCACHE_OBJLIST_CONFIG_WORK	0x00001000	/* show objects with work */
+#define FSCACHE_OBJLIST_CONFIG_NOWORK	0x00002000	/* show objects without work */
 
 	u8		buf[512];	/* key and aux data buffer */
 };
@@ -231,12 +231,11 @@ static int fscache_objlist_show(struct seq_file *m, void *v)
 		       READS, NOREADS);
 		FILTER(obj->events & obj->event_mask,
 		       EVENTS, NOEVENTS);
-		FILTER(obj->work.flags & ~(1UL << SLOW_WORK_VERY_SLOW),
-		       WORK, NOWORK);
+		FILTER(work_busy(&obj->work), WORK, NOWORK);
 	}
 
 	seq_printf(m,
-		   "%8x %8x %s %5u %3u %3u %3u %2u %5u %2lx %2lx %1lx %1lx | ",
+		   "%8x %8x %s %5u %3u %3u %3u %2u %5u %2lx %2lx %1lx %1x | ",
 		   obj->debug_id,
 		   obj->parent ? obj->parent->debug_id : -1,
 		   fscache_object_states_short[obj->state],
@@ -249,7 +248,7 @@ static int fscache_objlist_show(struct seq_file *m, void *v)
 		   obj->event_mask & FSCACHE_OBJECT_EVENTS_MASK,
 		   obj->events,
 		   obj->flags,
-		   obj->work.flags);
+		   work_busy(&obj->work));
 
 	no_cookie = true;
 	keylen = auxlen = 0;
diff --git a/fs/fscache/object.c b/fs/fscache/object.c
index 0b589a9b4ff..b6b897c550a 100644
--- a/fs/fscache/object.c
+++ b/fs/fscache/object.c
@@ -14,7 +14,6 @@
 
 #define FSCACHE_DEBUG_LEVEL COOKIE
 #include <linux/module.h>
-#include <linux/seq_file.h>
 #include "internal.h"
 
 const char *fscache_object_states[FSCACHE_OBJECT__NSTATES] = {
@@ -50,12 +49,8 @@ const char fscache_object_states_short[FSCACHE_OBJECT__NSTATES][5] = {
 	[FSCACHE_OBJECT_DEAD]		= "DEAD",
 };
 
-static void fscache_object_slow_work_put_ref(struct slow_work *);
-static int  fscache_object_slow_work_get_ref(struct slow_work *);
-static void fscache_object_slow_work_execute(struct slow_work *);
-#ifdef CONFIG_SLOW_WORK_DEBUG
-static void fscache_object_slow_work_desc(struct slow_work *, struct seq_file *);
-#endif
+static int  fscache_get_object(struct fscache_object *);
+static void fscache_put_object(struct fscache_object *);
 static void fscache_initialise_object(struct fscache_object *);
 static void fscache_lookup_object(struct fscache_object *);
 static void fscache_object_available(struct fscache_object *);
@@ -64,17 +59,6 @@ static void fscache_withdraw_object(struct fscache_object *);
 static void fscache_enqueue_dependents(struct fscache_object *);
 static void fscache_dequeue_object(struct fscache_object *);
 
-const struct slow_work_ops fscache_object_slow_work_ops = {
-	.owner		= THIS_MODULE,
-	.get_ref	= fscache_object_slow_work_get_ref,
-	.put_ref	= fscache_object_slow_work_put_ref,
-	.execute	= fscache_object_slow_work_execute,
-#ifdef CONFIG_SLOW_WORK_DEBUG
-	.desc		= fscache_object_slow_work_desc,
-#endif
-};
-EXPORT_SYMBOL(fscache_object_slow_work_ops);
-
 /*
  * we need to notify the parent when an op completes that we had outstanding
  * upon it
@@ -345,7 +329,7 @@ unsupported_event:
 /*
  * execute an object
  */
-static void fscache_object_slow_work_execute(struct slow_work *work)
+void fscache_object_work_func(struct work_struct *work)
 {
 	struct fscache_object *object =
 		container_of(work, struct fscache_object, work);
@@ -359,23 +343,9 @@ static void fscache_object_slow_work_execute(struct slow_work *work)
 	if (object->events & object->event_mask)
 		fscache_enqueue_object(object);
 	clear_bit(FSCACHE_OBJECT_EV_REQUEUE, &object->events);
+	fscache_put_object(object);
 }
-
-/*
- * describe an object for slow-work debugging
- */
-#ifdef CONFIG_SLOW_WORK_DEBUG
-static void fscache_object_slow_work_desc(struct slow_work *work,
-					  struct seq_file *m)
-{
-	struct fscache_object *object =
-		container_of(work, struct fscache_object, work);
-
-	seq_printf(m, "FSC: OBJ%x: %s",
-		   object->debug_id,
-		   fscache_object_states_short[object->state]);
-}
-#endif
+EXPORT_SYMBOL(fscache_object_work_func);
 
 /*
  * initialise an object
@@ -393,7 +363,6 @@ static void fscache_initialise_object(struct fscache_object *object)
 	_enter("");
 	ASSERT(object->cookie != NULL);
 	ASSERT(object->cookie->parent != NULL);
-	ASSERT(list_empty(&object->work.link));
 
 	if (object->events & ((1 << FSCACHE_OBJECT_EV_ERROR) |
 			      (1 << FSCACHE_OBJECT_EV_RELEASE) |
@@ -671,10 +640,8 @@ static void fscache_drop_object(struct fscache_object *object)
 		object->parent = NULL;
 	}
 
-	/* this just shifts the object release to the slow work processor */
-	fscache_stat(&fscache_n_cop_put_object);
-	object->cache->ops->put_object(object);
-	fscache_stat_d(&fscache_n_cop_put_object);
+	/* this just shifts the object release to the work processor */
+	fscache_put_object(object);
 
 	_leave("");
 }
@@ -758,12 +725,10 @@ void fscache_withdrawing_object(struct fscache_cache *cache,
 }
 
 /*
- * allow the slow work item processor to get a ref on an object
+ * get a ref on an object
  */
-static int fscache_object_slow_work_get_ref(struct slow_work *work)
+static int fscache_get_object(struct fscache_object *object)
 {
-	struct fscache_object *object =
-		container_of(work, struct fscache_object, work);
 	int ret;
 
 	fscache_stat(&fscache_n_cop_grab_object);
@@ -773,13 +738,10 @@ static int fscache_object_slow_work_get_ref(struct slow_work *work)
 }
 
 /*
- * allow the slow work item processor to discard a ref on a work item
+ * discard a ref on a work item
  */
-static void fscache_object_slow_work_put_ref(struct slow_work *work)
+static void fscache_put_object(struct fscache_object *object)
 {
-	struct fscache_object *object =
-		container_of(work, struct fscache_object, work);
-
 	fscache_stat(&fscache_n_cop_put_object);
 	object->cache->ops->put_object(object);
 	fscache_stat_d(&fscache_n_cop_put_object);
@@ -792,8 +754,48 @@ void fscache_enqueue_object(struct fscache_object *object)
 {
 	_enter("{OBJ%x}", object->debug_id);
 
-	slow_work_enqueue(&object->work);
+	if (fscache_get_object(object) >= 0) {
+		wait_queue_head_t *cong_wq =
+			&get_cpu_var(fscache_object_cong_wait);
+
+		if (queue_work(fscache_object_wq, &object->work)) {
+			if (fscache_object_congested())
+				wake_up(cong_wq);
+		} else
+			fscache_put_object(object);
+
+		put_cpu_var(fscache_object_cong_wait);
+	}
+}
+
+/**
+ * fscache_object_sleep_till_congested - Sleep until object wq is congested
+ * @timoutp: Scheduler sleep timeout
+ *
+ * Allow an object handler to sleep until the object workqueue is congested.
+ *
+ * The caller must set up a wake up event before calling this and must have set
+ * the appropriate sleep mode (such as TASK_UNINTERRUPTIBLE) and tested its own
+ * condition before calling this function as no test is made here.
+ *
+ * %true is returned if the object wq is congested, %false otherwise.
+ */
+bool fscache_object_sleep_till_congested(signed long *timeoutp)
+{
+	wait_queue_head_t *cong_wq = &__get_cpu_var(fscache_object_cong_wait);
+	DEFINE_WAIT(wait);
+
+	if (fscache_object_congested())
+		return true;
+
+	add_wait_queue_exclusive(cong_wq, &wait);
+	if (!fscache_object_congested())
+		*timeoutp = schedule_timeout(*timeoutp);
+	finish_wait(cong_wq, &wait);
+
+	return fscache_object_congested();
 }
+EXPORT_SYMBOL_GPL(fscache_object_sleep_till_congested);
 
 /*
  * enqueue the dependents of an object for metadata-type processing
@@ -819,9 +821,7 @@ static void fscache_enqueue_dependents(struct fscache_object *object)
 
 		/* sort onto appropriate lists */
 		fscache_enqueue_object(dep);
-		fscache_stat(&fscache_n_cop_put_object);
-		dep->cache->ops->put_object(dep);
-		fscache_stat_d(&fscache_n_cop_put_object);
+		fscache_put_object(dep);
 
 		if (!list_empty(&object->dependents))
 			cond_resched_lock(&object->lock);
diff --git a/fs/fscache/operation.c b/fs/fscache/operation.c
index f17cecafae4..b9f34eaede0 100644
--- a/fs/fscache/operation.c
+++ b/fs/fscache/operation.c
@@ -42,16 +42,12 @@ void fscache_enqueue_operation(struct fscache_operation *op)
 
 	fscache_stat(&fscache_n_op_enqueue);
 	switch (op->flags & FSCACHE_OP_TYPE) {
-	case FSCACHE_OP_FAST:
-		_debug("queue fast");
+	case FSCACHE_OP_ASYNC:
+		_debug("queue async");
 		atomic_inc(&op->usage);
-		if (!schedule_work(&op->fast_work))
+		if (!queue_work(fscache_op_wq, &op->work))
 			fscache_put_operation(op);
 		break;
-	case FSCACHE_OP_SLOW:
-		_debug("queue slow");
-		slow_work_enqueue(&op->slow_work);
-		break;
 	case FSCACHE_OP_MYTHREAD:
 		_debug("queue for caller's attention");
 		break;
@@ -455,36 +451,13 @@ void fscache_operation_gc(struct work_struct *work)
 }
 
 /*
- * allow the slow work item processor to get a ref on an operation
- */
-static int fscache_op_get_ref(struct slow_work *work)
-{
-	struct fscache_operation *op =
-		container_of(work, struct fscache_operation, slow_work);
-
-	atomic_inc(&op->usage);
-	return 0;
-}
-
-/*
- * allow the slow work item processor to discard a ref on an operation
- */
-static void fscache_op_put_ref(struct slow_work *work)
-{
-	struct fscache_operation *op =
-		container_of(work, struct fscache_operation, slow_work);
-
-	fscache_put_operation(op);
-}
-
-/*
- * execute an operation using the slow thread pool to provide processing context
- * - the caller holds a ref to this object, so we don't need to hold one
+ * execute an operation using fs_op_wq to provide processing context -
+ * the caller holds a ref to this object, so we don't need to hold one
  */
-static void fscache_op_execute(struct slow_work *work)
+void fscache_op_work_func(struct work_struct *work)
 {
 	struct fscache_operation *op =
-		container_of(work, struct fscache_operation, slow_work);
+		container_of(work, struct fscache_operation, work);
 	unsigned long start;
 
 	_enter("{OBJ%x OP%x,%d}",
@@ -494,31 +467,7 @@ static void fscache_op_execute(struct slow_work *work)
 	start = jiffies;
 	op->processor(op);
 	fscache_hist(fscache_ops_histogram, start);
+	fscache_put_operation(op);
 
 	_leave("");
 }
-
-/*
- * describe an operation for slow-work debugging
- */
-#ifdef CONFIG_SLOW_WORK_DEBUG
-static void fscache_op_desc(struct slow_work *work, struct seq_file *m)
-{
-	struct fscache_operation *op =
-		container_of(work, struct fscache_operation, slow_work);
-
-	seq_printf(m, "FSC: OBJ%x OP%x: %s/%s fl=%lx",
-		   op->object->debug_id, op->debug_id,
-		   op->name, op->state, op->flags);
-}
-#endif
-
-const struct slow_work_ops fscache_op_slow_work_ops = {
-	.owner		= THIS_MODULE,
-	.get_ref	= fscache_op_get_ref,
-	.put_ref	= fscache_op_put_ref,
-	.execute	= fscache_op_execute,
-#ifdef CONFIG_SLOW_WORK_DEBUG
-	.desc		= fscache_op_desc,
-#endif
-};
diff --git a/fs/fscache/page.c b/fs/fscache/page.c
index 723b889fd21..41c441c2058 100644
--- a/fs/fscache/page.c
+++ b/fs/fscache/page.c
@@ -105,7 +105,7 @@ bool __fscache_maybe_release_page(struct fscache_cookie *cookie,
 
 page_busy:
 	/* we might want to wait here, but that could deadlock the allocator as
-	 * the slow-work threads writing to the cache may all end up sleeping
+	 * the work threads writing to the cache may all end up sleeping
 	 * on memory allocation */
 	fscache_stat(&fscache_n_store_vmscan_busy);
 	return false;
@@ -188,9 +188,8 @@ int __fscache_attr_changed(struct fscache_cookie *cookie)
 		return -ENOMEM;
 	}
 
-	fscache_operation_init(op, NULL);
-	fscache_operation_init_slow(op, fscache_attr_changed_op);
-	op->flags = FSCACHE_OP_SLOW | (1 << FSCACHE_OP_EXCLUSIVE);
+	fscache_operation_init(op, fscache_attr_changed_op, NULL);
+	op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE);
 	fscache_set_op_name(op, "Attr");
 
 	spin_lock(&cookie->lock);
@@ -218,24 +217,6 @@ nobufs:
 EXPORT_SYMBOL(__fscache_attr_changed);
 
 /*
- * handle secondary execution given to a retrieval op on behalf of the
- * cache
- */
-static void fscache_retrieval_work(struct work_struct *work)
-{
-	struct fscache_retrieval *op =
-		container_of(work, struct fscache_retrieval, op.fast_work);
-	unsigned long start;
-
-	_enter("{OP%x}", op->op.debug_id);
-
-	start = jiffies;
-	op->op.processor(&op->op);
-	fscache_hist(fscache_ops_histogram, start);
-	fscache_put_operation(&op->op);
-}
-
-/*
  * release a retrieval op reference
  */
 static void fscache_release_retrieval_op(struct fscache_operation *_op)
@@ -269,13 +250,12 @@ static struct fscache_retrieval *fscache_alloc_retrieval(
 		return NULL;
 	}
 
-	fscache_operation_init(&op->op, fscache_release_retrieval_op);
+	fscache_operation_init(&op->op, NULL, fscache_release_retrieval_op);
 	op->op.flags	= FSCACHE_OP_MYTHREAD | (1 << FSCACHE_OP_WAITING);
 	op->mapping	= mapping;
 	op->end_io_func	= end_io_func;
 	op->context	= context;
 	op->start_time	= jiffies;
-	INIT_WORK(&op->op.fast_work, fscache_retrieval_work);
 	INIT_LIST_HEAD(&op->to_do);
 	fscache_set_op_name(&op->op, "Retr");
 	return op;
@@ -795,9 +775,9 @@ int __fscache_write_page(struct fscache_cookie *cookie,
 	if (!op)
 		goto nomem;
 
-	fscache_operation_init(&op->op, fscache_release_write_op);
-	fscache_operation_init_slow(&op->op, fscache_write_op);
-	op->op.flags = FSCACHE_OP_SLOW | (1 << FSCACHE_OP_WAITING);
+	fscache_operation_init(&op->op, fscache_write_op,
+			       fscache_release_write_op);
+	op->op.flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_WAITING);
 	fscache_set_op_name(&op->op, "Write1");
 
 	ret = radix_tree_preload(gfp & ~__GFP_HIGHMEM);
@@ -852,7 +832,7 @@ int __fscache_write_page(struct fscache_cookie *cookie,
 	fscache_stat(&fscache_n_store_ops);
 	fscache_stat(&fscache_n_stores_ok);
 
-	/* the slow work queue now carries its own ref on the object */
+	/* the work queue now carries its own ref on the object */
 	fscache_put_operation(&op->op);
 	_leave(" = 0");
 	return 0;
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 9424796d663..69ad053ffd7 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -239,7 +239,6 @@ static u64 fuse_get_unique(struct fuse_conn *fc)
 
 static void queue_request(struct fuse_conn *fc, struct fuse_req *req)
 {
-	req->in.h.unique = fuse_get_unique(fc);
 	req->in.h.len = sizeof(struct fuse_in_header) +
 		len_args(req->in.numargs, (struct fuse_arg *) req->in.args);
 	list_add_tail(&req->list, &fc->pending);
@@ -261,6 +260,7 @@ static void flush_bg_queue(struct fuse_conn *fc)
 		req = list_entry(fc->bg_queue.next, struct fuse_req, list);
 		list_del(&req->list);
 		fc->active_background++;
+		req->in.h.unique = fuse_get_unique(fc);
 		queue_request(fc, req);
 	}
 }
@@ -398,6 +398,7 @@ void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req)
 	else if (fc->conn_error)
 		req->out.h.error = -ECONNREFUSED;
 	else {
+		req->in.h.unique = fuse_get_unique(fc);
 		queue_request(fc, req);
 		/* acquire extra reference, since request is still needed
 		   after request_end() */
@@ -450,6 +451,23 @@ void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req)
 }
 EXPORT_SYMBOL_GPL(fuse_request_send_background);
 
+static int fuse_request_send_notify_reply(struct fuse_conn *fc,
+					  struct fuse_req *req, u64 unique)
+{
+	int err = -ENODEV;
+
+	req->isreply = 0;
+	req->in.h.unique = unique;
+	spin_lock(&fc->lock);
+	if (fc->connected) {
+		queue_request(fc, req);
+		err = 0;
+	}
+	spin_unlock(&fc->lock);
+
+	return err;
+}
+
 /*
  * Called under fc->lock
  *
@@ -535,13 +553,13 @@ static void fuse_copy_finish(struct fuse_copy_state *cs)
 		if (!cs->write) {
 			buf->ops->unmap(cs->pipe, buf, cs->mapaddr);
 		} else {
-			kunmap_atomic(cs->mapaddr, KM_USER0);
+			kunmap(buf->page);
 			buf->len = PAGE_SIZE - cs->len;
 		}
 		cs->currbuf = NULL;
 		cs->mapaddr = NULL;
 	} else if (cs->mapaddr) {
-		kunmap_atomic(cs->mapaddr, KM_USER0);
+		kunmap(cs->pg);
 		if (cs->write) {
 			flush_dcache_page(cs->pg);
 			set_page_dirty_lock(cs->pg);
@@ -572,7 +590,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs)
 
 			BUG_ON(!cs->nr_segs);
 			cs->currbuf = buf;
-			cs->mapaddr = buf->ops->map(cs->pipe, buf, 1);
+			cs->mapaddr = buf->ops->map(cs->pipe, buf, 0);
 			cs->len = buf->len;
 			cs->buf = cs->mapaddr + buf->offset;
 			cs->pipebufs++;
@@ -592,7 +610,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs)
 			buf->len = 0;
 
 			cs->currbuf = buf;
-			cs->mapaddr = kmap_atomic(page, KM_USER0);
+			cs->mapaddr = kmap(page);
 			cs->buf = cs->mapaddr;
 			cs->len = PAGE_SIZE;
 			cs->pipebufs++;
@@ -611,7 +629,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs)
 			return err;
 		BUG_ON(err != 1);
 		offset = cs->addr % PAGE_SIZE;
-		cs->mapaddr = kmap_atomic(cs->pg, KM_USER0);
+		cs->mapaddr = kmap(cs->pg);
 		cs->buf = cs->mapaddr + offset;
 		cs->len = min(PAGE_SIZE - offset, cs->seglen);
 		cs->seglen -= cs->len;
@@ -1231,6 +1249,199 @@ err:
 	return err;
 }
 
+static int fuse_notify_store(struct fuse_conn *fc, unsigned int size,
+			     struct fuse_copy_state *cs)
+{
+	struct fuse_notify_store_out outarg;
+	struct inode *inode;
+	struct address_space *mapping;
+	u64 nodeid;
+	int err;
+	pgoff_t index;
+	unsigned int offset;
+	unsigned int num;
+	loff_t file_size;
+	loff_t end;
+
+	err = -EINVAL;
+	if (size < sizeof(outarg))
+		goto out_finish;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		goto out_finish;
+
+	err = -EINVAL;
+	if (size - sizeof(outarg) != outarg.size)
+		goto out_finish;
+
+	nodeid = outarg.nodeid;
+
+	down_read(&fc->killsb);
+
+	err = -ENOENT;
+	if (!fc->sb)
+		goto out_up_killsb;
+
+	inode = ilookup5(fc->sb, nodeid, fuse_inode_eq, &nodeid);
+	if (!inode)
+		goto out_up_killsb;
+
+	mapping = inode->i_mapping;
+	index = outarg.offset >> PAGE_CACHE_SHIFT;
+	offset = outarg.offset & ~PAGE_CACHE_MASK;
+	file_size = i_size_read(inode);
+	end = outarg.offset + outarg.size;
+	if (end > file_size) {
+		file_size = end;
+		fuse_write_update_size(inode, file_size);
+	}
+
+	num = outarg.size;
+	while (num) {
+		struct page *page;
+		unsigned int this_num;
+
+		err = -ENOMEM;
+		page = find_or_create_page(mapping, index,
+					   mapping_gfp_mask(mapping));
+		if (!page)
+			goto out_iput;
+
+		this_num = min_t(unsigned, num, PAGE_CACHE_SIZE - offset);
+		err = fuse_copy_page(cs, &page, offset, this_num, 0);
+		if (!err && offset == 0 && (num != 0 || file_size == end))
+			SetPageUptodate(page);
+		unlock_page(page);
+		page_cache_release(page);
+
+		if (err)
+			goto out_iput;
+
+		num -= this_num;
+		offset = 0;
+		index++;
+	}
+
+	err = 0;
+
+out_iput:
+	iput(inode);
+out_up_killsb:
+	up_read(&fc->killsb);
+out_finish:
+	fuse_copy_finish(cs);
+	return err;
+}
+
+static void fuse_retrieve_end(struct fuse_conn *fc, struct fuse_req *req)
+{
+	int i;
+
+	for (i = 0; i < req->num_pages; i++) {
+		struct page *page = req->pages[i];
+		page_cache_release(page);
+	}
+}
+
+static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode,
+			 struct fuse_notify_retrieve_out *outarg)
+{
+	int err;
+	struct address_space *mapping = inode->i_mapping;
+	struct fuse_req *req;
+	pgoff_t index;
+	loff_t file_size;
+	unsigned int num;
+	unsigned int offset;
+	size_t total_len;
+
+	req = fuse_get_req(fc);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+
+	offset = outarg->offset & ~PAGE_CACHE_MASK;
+
+	req->in.h.opcode = FUSE_NOTIFY_REPLY;
+	req->in.h.nodeid = outarg->nodeid;
+	req->in.numargs = 2;
+	req->in.argpages = 1;
+	req->page_offset = offset;
+	req->end = fuse_retrieve_end;
+
+	index = outarg->offset >> PAGE_CACHE_SHIFT;
+	file_size = i_size_read(inode);
+	num = outarg->size;
+	if (outarg->offset > file_size)
+		num = 0;
+	else if (outarg->offset + num > file_size)
+		num = file_size - outarg->offset;
+
+	while (num) {
+		struct page *page;
+		unsigned int this_num;
+
+		page = find_get_page(mapping, index);
+		if (!page)
+			break;
+
+		this_num = min_t(unsigned, num, PAGE_CACHE_SIZE - offset);
+		req->pages[req->num_pages] = page;
+		req->num_pages++;
+
+		num -= this_num;
+		total_len += this_num;
+	}
+	req->misc.retrieve_in.offset = outarg->offset;
+	req->misc.retrieve_in.size = total_len;
+	req->in.args[0].size = sizeof(req->misc.retrieve_in);
+	req->in.args[0].value = &req->misc.retrieve_in;
+	req->in.args[1].size = total_len;
+
+	err = fuse_request_send_notify_reply(fc, req, outarg->notify_unique);
+	if (err)
+		fuse_retrieve_end(fc, req);
+
+	return err;
+}
+
+static int fuse_notify_retrieve(struct fuse_conn *fc, unsigned int size,
+				struct fuse_copy_state *cs)
+{
+	struct fuse_notify_retrieve_out outarg;
+	struct inode *inode;
+	int err;
+
+	err = -EINVAL;
+	if (size != sizeof(outarg))
+		goto copy_finish;
+
+	err = fuse_copy_one(cs, &outarg, sizeof(outarg));
+	if (err)
+		goto copy_finish;
+
+	fuse_copy_finish(cs);
+
+	down_read(&fc->killsb);
+	err = -ENOENT;
+	if (fc->sb) {
+		u64 nodeid = outarg.nodeid;
+
+		inode = ilookup5(fc->sb, nodeid, fuse_inode_eq, &nodeid);
+		if (inode) {
+			err = fuse_retrieve(fc, inode, &outarg);
+			iput(inode);
+		}
+	}
+	up_read(&fc->killsb);
+
+	return err;
+
+copy_finish:
+	fuse_copy_finish(cs);
+	return err;
+}
+
 static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code,
 		       unsigned int size, struct fuse_copy_state *cs)
 {
@@ -1244,6 +1455,12 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code,
 	case FUSE_NOTIFY_INVAL_ENTRY:
 		return fuse_notify_inval_entry(fc, size, cs);
 
+	case FUSE_NOTIFY_STORE:
+		return fuse_notify_store(fc, size, cs);
+
+	case FUSE_NOTIFY_RETRIEVE:
+		return fuse_notify_retrieve(fc, size, cs);
+
 	default:
 		fuse_copy_finish(cs);
 		return -EINVAL;
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index ada0adeb3bb..147c1f71bdb 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -706,7 +706,7 @@ static int fuse_write_begin(struct file *file, struct address_space *mapping,
 	return 0;
 }
 
-static void fuse_write_update_size(struct inode *inode, loff_t pos)
+void fuse_write_update_size(struct inode *inode, loff_t pos)
 {
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	struct fuse_inode *fi = get_fuse_inode(inode);
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index 8f309f04064..57d4a3a0f10 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -272,6 +272,7 @@ struct fuse_req {
 			struct fuse_write_in in;
 			struct fuse_write_out out;
 		} write;
+		struct fuse_notify_retrieve_in retrieve_in;
 		struct fuse_lk_in lk_in;
 	} misc;
 
@@ -748,4 +749,6 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg,
 unsigned fuse_file_poll(struct file *file, poll_table *wait);
 int fuse_dev_release(struct inode *inode, struct file *file);
 
+void fuse_write_update_size(struct inode *inode, loff_t pos);
+
 #endif /* _FS_FUSE_I_H */
diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig
index a47b4310711..cc966552214 100644
--- a/fs/gfs2/Kconfig
+++ b/fs/gfs2/Kconfig
@@ -7,7 +7,6 @@ config GFS2_FS
 	select IP_SCTP if DLM_SCTP
 	select FS_POSIX_ACL
 	select CRC32
-	select SLOW_WORK
 	select QUOTACTL
 	help
 	  A cluster filesystem.
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 8fcbce48a12..fdbf4b366fa 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -12,7 +12,6 @@
 
 #include <linux/fs.h>
 #include <linux/workqueue.h>
-#include <linux/slow-work.h>
 #include <linux/dlm.h>
 #include <linux/buffer_head.h>
 
@@ -383,7 +382,7 @@ struct gfs2_journal_extent {
 struct gfs2_jdesc {
 	struct list_head jd_list;
 	struct list_head extent_list;
-	struct slow_work jd_work;
+	struct work_struct jd_work;
 	struct inode *jd_inode;
 	unsigned long jd_flags;
 #define JDF_RECOVERY 1
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index fb2a5f93b7c..b1e9630eb46 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -15,7 +15,6 @@
 #include <linux/init.h>
 #include <linux/gfs2_ondisk.h>
 #include <asm/atomic.h>
-#include <linux/slow-work.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -24,6 +23,7 @@
 #include "util.h"
 #include "glock.h"
 #include "quota.h"
+#include "recovery.h"
 
 static struct shrinker qd_shrinker = {
 	.shrink = gfs2_shrink_qd_memory,
@@ -138,9 +138,11 @@ static int __init init_gfs2_fs(void)
 	if (error)
 		goto fail_unregister;
 
-	error = slow_work_register_user(THIS_MODULE);
-	if (error)
-		goto fail_slow;
+	error = -ENOMEM;
+	gfs_recovery_wq = alloc_workqueue("gfs_recovery",
+					  WQ_NON_REENTRANT | WQ_RESCUER, 0);
+	if (!gfs_recovery_wq)
+		goto fail_wq;
 
 	gfs2_register_debugfs();
 
@@ -148,7 +150,7 @@ static int __init init_gfs2_fs(void)
 
 	return 0;
 
-fail_slow:
+fail_wq:
 	unregister_filesystem(&gfs2meta_fs_type);
 fail_unregister:
 	unregister_filesystem(&gfs2_fs_type);
@@ -190,7 +192,7 @@ static void __exit exit_gfs2_fs(void)
 	gfs2_unregister_debugfs();
 	unregister_filesystem(&gfs2_fs_type);
 	unregister_filesystem(&gfs2meta_fs_type);
-	slow_work_unregister_user(THIS_MODULE);
+	destroy_workqueue(gfs_recovery_wq);
 
 	kmem_cache_destroy(gfs2_quotad_cachep);
 	kmem_cache_destroy(gfs2_rgrpd_cachep);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 45a4a36195d..4f44bdeb2f0 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -17,7 +17,6 @@
 #include <linux/namei.h>
 #include <linux/mount.h>
 #include <linux/gfs2_ondisk.h>
-#include <linux/slow-work.h>
 #include <linux/quotaops.h>
 
 #include "gfs2.h"
@@ -673,7 +672,7 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
 			break;
 
 		INIT_LIST_HEAD(&jd->extent_list);
-		slow_work_init(&jd->jd_work, &gfs2_recover_ops);
+		INIT_WORK(&jd->jd_work, gfs2_recover_func);
 		jd->jd_inode = gfs2_lookupi(sdp->sd_jindex, &name, 1);
 		if (!jd->jd_inode || IS_ERR(jd->jd_inode)) {
 			if (!jd->jd_inode)
@@ -782,7 +781,8 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 	if (sdp->sd_lockstruct.ls_first) {
 		unsigned int x;
 		for (x = 0; x < sdp->sd_journals; x++) {
-			error = gfs2_recover_journal(gfs2_jdesc_find(sdp, x));
+			error = gfs2_recover_journal(gfs2_jdesc_find(sdp, x),
+						     true);
 			if (error) {
 				fs_err(sdp, "error recovering journal %u: %d\n",
 				       x, error);
@@ -792,7 +792,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 
 		gfs2_others_may_mount(sdp);
 	} else if (!sdp->sd_args.ar_spectator) {
-		error = gfs2_recover_journal(sdp->sd_jdesc);
+		error = gfs2_recover_journal(sdp->sd_jdesc, true);
 		if (error) {
 			fs_err(sdp, "error recovering my journal: %d\n", error);
 			goto fail_jinode_gh;
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index 8bb643cb265..1bc6b5695e6 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -1449,10 +1449,10 @@ static int gfs2_quota_get_xstate(struct super_block *sb,
 
 	switch (sdp->sd_args.ar_quota) {
 	case GFS2_QUOTA_ON:
-		fqs->qs_flags |= (XFS_QUOTA_UDQ_ENFD | XFS_QUOTA_GDQ_ENFD);
+		fqs->qs_flags |= (FS_QUOTA_UDQ_ENFD | FS_QUOTA_GDQ_ENFD);
 		/*FALLTHRU*/
 	case GFS2_QUOTA_ACCOUNT:
-		fqs->qs_flags |= (XFS_QUOTA_UDQ_ACCT | XFS_QUOTA_GDQ_ACCT);
+		fqs->qs_flags |= (FS_QUOTA_UDQ_ACCT | FS_QUOTA_GDQ_ACCT);
 		break;
 	case GFS2_QUOTA_OFF:
 		break;
@@ -1498,7 +1498,7 @@ static int gfs2_get_dqblk(struct super_block *sb, int type, qid_t id,
 
 	qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb;
 	fdq->d_version = FS_DQUOT_VERSION;
-	fdq->d_flags = (type == QUOTA_USER) ? XFS_USER_QUOTA : XFS_GROUP_QUOTA;
+	fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA;
 	fdq->d_id = id;
 	fdq->d_blk_hardlimit = be64_to_cpu(qlvb->qb_limit);
 	fdq->d_blk_softlimit = be64_to_cpu(qlvb->qb_warn);
@@ -1533,12 +1533,12 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id,
 	switch(type) {
 	case USRQUOTA:
 		type = QUOTA_USER;
-		if (fdq->d_flags != XFS_USER_QUOTA)
+		if (fdq->d_flags != FS_USER_QUOTA)
 			return -EINVAL;
 		break;
 	case GRPQUOTA:
 		type = QUOTA_GROUP;
-		if (fdq->d_flags != XFS_GROUP_QUOTA)
+		if (fdq->d_flags != FS_GROUP_QUOTA)
 			return -EINVAL;
 		break;
 	default:
diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c
index 4b9bece3d43..f7f89a94a5a 100644
--- a/fs/gfs2/recovery.c
+++ b/fs/gfs2/recovery.c
@@ -14,7 +14,6 @@
 #include <linux/buffer_head.h>
 #include <linux/gfs2_ondisk.h>
 #include <linux/crc32.h>
-#include <linux/slow-work.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -28,6 +27,8 @@
 #include "util.h"
 #include "dir.h"
 
+struct workqueue_struct *gfs_recovery_wq;
+
 int gfs2_replay_read_block(struct gfs2_jdesc *jd, unsigned int blk,
 			   struct buffer_head **bh)
 {
@@ -443,23 +444,7 @@ static void gfs2_recovery_done(struct gfs2_sbd *sdp, unsigned int jid,
         kobject_uevent_env(&sdp->sd_kobj, KOBJ_CHANGE, envp);
 }
 
-static int gfs2_recover_get_ref(struct slow_work *work)
-{
-	struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work);
-	if (test_and_set_bit(JDF_RECOVERY, &jd->jd_flags))
-		return -EBUSY;
-	return 0;
-}
-
-static void gfs2_recover_put_ref(struct slow_work *work)
-{
-	struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work);
-	clear_bit(JDF_RECOVERY, &jd->jd_flags);
-	smp_mb__after_clear_bit();
-	wake_up_bit(&jd->jd_flags, JDF_RECOVERY);
-}
-
-static void gfs2_recover_work(struct slow_work *work)
+void gfs2_recover_func(struct work_struct *work)
 {
 	struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work);
 	struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
@@ -578,7 +563,7 @@ static void gfs2_recover_work(struct slow_work *work)
 		gfs2_glock_dq_uninit(&j_gh);
 
 	fs_info(sdp, "jid=%u: Done\n", jd->jd_jid);
-	return;
+	goto done;
 
 fail_gunlock_tr:
 	gfs2_glock_dq_uninit(&t_gh);
@@ -590,32 +575,35 @@ fail_gunlock_j:
 	}
 
 	fs_info(sdp, "jid=%u: %s\n", jd->jd_jid, (error) ? "Failed" : "Done");
-
 fail:
 	gfs2_recovery_done(sdp, jd->jd_jid, LM_RD_GAVEUP);
+done:
+	clear_bit(JDF_RECOVERY, &jd->jd_flags);
+	smp_mb__after_clear_bit();
+	wake_up_bit(&jd->jd_flags, JDF_RECOVERY);
 }
 
-struct slow_work_ops gfs2_recover_ops = {
-	.owner	 = THIS_MODULE,
-	.get_ref = gfs2_recover_get_ref,
-	.put_ref = gfs2_recover_put_ref,
-	.execute = gfs2_recover_work,
-};
-
-
 static int gfs2_recovery_wait(void *word)
 {
 	schedule();
 	return 0;
 }
 
-int gfs2_recover_journal(struct gfs2_jdesc *jd)
+int gfs2_recover_journal(struct gfs2_jdesc *jd, bool wait)
 {
 	int rv;
-	rv = slow_work_enqueue(&jd->jd_work);
-	if (rv)
-		return rv;
-	wait_on_bit(&jd->jd_flags, JDF_RECOVERY, gfs2_recovery_wait, TASK_UNINTERRUPTIBLE);
+
+	if (test_and_set_bit(JDF_RECOVERY, &jd->jd_flags))
+		return -EBUSY;
+
+	/* we have JDF_RECOVERY, queue should always succeed */
+	rv = queue_work(gfs_recovery_wq, &jd->jd_work);
+	BUG_ON(!rv);
+
+	if (wait)
+		wait_on_bit(&jd->jd_flags, JDF_RECOVERY, gfs2_recovery_wait,
+			    TASK_UNINTERRUPTIBLE);
+
 	return 0;
 }
 
diff --git a/fs/gfs2/recovery.h b/fs/gfs2/recovery.h
index 1616ac22569..2226136c764 100644
--- a/fs/gfs2/recovery.h
+++ b/fs/gfs2/recovery.h
@@ -12,6 +12,8 @@
 
 #include "incore.h"
 
+extern struct workqueue_struct *gfs_recovery_wq;
+
 static inline void gfs2_replay_incr_blk(struct gfs2_sbd *sdp, unsigned int *blk)
 {
 	if (++*blk == sdp->sd_jdesc->jd_blocks)
@@ -27,8 +29,8 @@ extern void gfs2_revoke_clean(struct gfs2_sbd *sdp);
 
 extern int gfs2_find_jhead(struct gfs2_jdesc *jd,
 		    struct gfs2_log_header_host *head);
-extern int gfs2_recover_journal(struct gfs2_jdesc *gfs2_jd);
-extern struct slow_work_ops gfs2_recover_ops;
+extern int gfs2_recover_journal(struct gfs2_jdesc *gfs2_jd, bool wait);
+extern void gfs2_recover_func(struct work_struct *work);
 
 #endif /* __RECOVERY_DOT_H__ */
 
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index d019d0d55e0..ccacffd2faa 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -25,6 +25,7 @@
 #include "quota.h"
 #include "util.h"
 #include "glops.h"
+#include "recovery.h"
 
 struct gfs2_attr {
 	struct attribute attr;
@@ -376,7 +377,7 @@ static ssize_t recover_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
 	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
 		if (jd->jd_jid != jid)
 			continue;
-		rv = slow_work_enqueue(&jd->jd_work);
+		rv = gfs2_recover_journal(jd, false);
 		break;
 	}
 out:
diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c
index 93d1e47647b..f19ce94693d 100644
--- a/fs/jbd/journal.c
+++ b/fs/jbd/journal.c
@@ -1281,13 +1281,9 @@ int journal_check_used_features (journal_t *journal, unsigned long compat,
 int journal_check_available_features (journal_t *journal, unsigned long compat,
 				      unsigned long ro, unsigned long incompat)
 {
-	journal_superblock_t *sb;
-
 	if (!compat && !ro && !incompat)
 		return 1;
 
-	sb = journal->j_superblock;
-
 	/* We can support any known requested features iff the
 	 * superblock is in version 2.  Otherwise we fail to support any
 	 * extended sb features. */
@@ -1481,7 +1477,6 @@ int journal_flush(journal_t *journal)
 
 int journal_wipe(journal_t *journal, int write)
 {
-	journal_superblock_t *sb;
 	int err = 0;
 
 	J_ASSERT (!(journal->j_flags & JFS_LOADED));
@@ -1490,8 +1485,6 @@ int journal_wipe(journal_t *journal, int write)
 	if (err)
 		return err;
 
-	sb = journal->j_superblock;
-
 	if (!journal->j_tail)
 		goto no_recovery;
 
diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c
index 54c9bc9e1b1..81051dafebf 100644
--- a/fs/jbd/recovery.c
+++ b/fs/jbd/recovery.c
@@ -283,12 +283,9 @@ int journal_recover(journal_t *journal)
 int journal_skip_recovery(journal_t *journal)
 {
 	int			err;
-	journal_superblock_t *	sb;
-
 	struct recovery_info	info;
 
 	memset (&info, 0, sizeof(info));
-	sb = journal->j_superblock;
 
 	err = do_one_pass(journal, &info, PASS_SCAN);
 
@@ -297,7 +294,8 @@ int journal_skip_recovery(journal_t *journal)
 		++journal->j_transaction_sequence;
 	} else {
 #ifdef CONFIG_JBD_DEBUG
-		int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence);
+		int dropped = info.end_transaction -
+			      be32_to_cpu(journal->j_superblock->s_sequence);
 #endif
 		jbd_debug(1,
 			  "JBD: ignoring %d transaction%s from the journal.\n",
@@ -321,11 +319,6 @@ static int do_one_pass(journal_t *journal,
 	unsigned int		sequence;
 	int			blocktype;
 
-	/* Precompute the maximum metadata descriptors in a descriptor block */
-	int			MAX_BLOCKS_PER_DESC;
-	MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t))
-			       / sizeof(journal_block_tag_t));
-
 	/*
 	 * First thing is to establish what we expect to find in the log
 	 * (in terms of transaction IDs), and where (in terms of log
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 076d1cc44f9..1c23a0f4e8a 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -118,13 +118,13 @@ static int __try_to_free_cp_buf(struct journal_head *jh)
 void __jbd2_log_wait_for_space(journal_t *journal)
 {
 	int nblocks, space_left;
-	assert_spin_locked(&journal->j_state_lock);
+	/* assert_spin_locked(&journal->j_state_lock); */
 
 	nblocks = jbd_space_needed(journal);
 	while (__jbd2_log_space_left(journal) < nblocks) {
 		if (journal->j_flags & JBD2_ABORT)
 			return;
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		mutex_lock(&journal->j_checkpoint_mutex);
 
 		/*
@@ -138,7 +138,7 @@ void __jbd2_log_wait_for_space(journal_t *journal)
 		 * filesystem, so abort the journal and leave a stack
 		 * trace for forensic evidence.
 		 */
-		spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
 		spin_lock(&journal->j_list_lock);
 		nblocks = jbd_space_needed(journal);
 		space_left = __jbd2_log_space_left(journal);
@@ -149,7 +149,7 @@ void __jbd2_log_wait_for_space(journal_t *journal)
 			if (journal->j_committing_transaction)
 				tid = journal->j_committing_transaction->t_tid;
 			spin_unlock(&journal->j_list_lock);
-			spin_unlock(&journal->j_state_lock);
+			write_unlock(&journal->j_state_lock);
 			if (chkpt) {
 				jbd2_log_do_checkpoint(journal);
 			} else if (jbd2_cleanup_journal_tail(journal) == 0) {
@@ -167,7 +167,7 @@ void __jbd2_log_wait_for_space(journal_t *journal)
 				WARN_ON(1);
 				jbd2_journal_abort(journal, 0);
 			}
-			spin_lock(&journal->j_state_lock);
+			write_lock(&journal->j_state_lock);
 		} else {
 			spin_unlock(&journal->j_list_lock);
 		}
@@ -474,7 +474,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 	 * next transaction ID we will write, and where it will
 	 * start. */
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	spin_lock(&journal->j_list_lock);
 	transaction = journal->j_checkpoint_transactions;
 	if (transaction) {
@@ -496,7 +496,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 	/* If the oldest pinned transaction is at the tail of the log
            already then there's not much we can do right now. */
 	if (journal->j_tail_sequence == first_tid) {
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		return 1;
 	}
 
@@ -516,7 +516,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 	journal->j_free += freed;
 	journal->j_tail_sequence = first_tid;
 	journal->j_tail = blocknr;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 
 	/*
 	 * If there is an external journal, we need to make sure that
@@ -775,7 +775,7 @@ void __jbd2_journal_drop_transaction(journal_t *journal, transaction_t *transact
 	J_ASSERT(transaction->t_log_list == NULL);
 	J_ASSERT(transaction->t_checkpoint_list == NULL);
 	J_ASSERT(transaction->t_checkpoint_io_list == NULL);
-	J_ASSERT(transaction->t_updates == 0);
+	J_ASSERT(atomic_read(&transaction->t_updates) == 0);
 	J_ASSERT(journal->j_committing_transaction != transaction);
 	J_ASSERT(journal->j_running_transaction != transaction);
 
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 75716d3d2be..f52e5e8049f 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -150,11 +150,11 @@ static int journal_submit_commit_record(journal_t *journal,
 	 */
 	if (ret == -EOPNOTSUPP && barrier_done) {
 		printk(KERN_WARNING
-		       "JBD: barrier-based sync failed on %s - "
-		       "disabling barriers\n", journal->j_devname);
-		spin_lock(&journal->j_state_lock);
+		       "JBD2: Disabling barriers on %s, "
+		       "not supported by device\n", journal->j_devname);
+		write_lock(&journal->j_state_lock);
 		journal->j_flags &= ~JBD2_BARRIER;
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 
 		/* And try again, without the barrier */
 		lock_buffer(bh);
@@ -180,11 +180,11 @@ retry:
 	wait_on_buffer(bh);
 	if (buffer_eopnotsupp(bh) && (journal->j_flags & JBD2_BARRIER)) {
 		printk(KERN_WARNING
-		       "JBD2: wait_on_commit_record: sync failed on %s - "
-		       "disabling barriers\n", journal->j_devname);
-		spin_lock(&journal->j_state_lock);
+		       "JBD2: %s: disabling barries on %s - not supported "
+		       "by device\n", __func__, journal->j_devname);
+		write_lock(&journal->j_state_lock);
 		journal->j_flags &= ~JBD2_BARRIER;
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 
 		lock_buffer(bh);
 		clear_buffer_dirty(bh);
@@ -400,7 +400,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 	jbd_debug(1, "JBD: starting commit of transaction %d\n",
 			commit_transaction->t_tid);
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_LOCKED;
 
 	/*
@@ -417,23 +417,23 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 					      stats.run.rs_locked);
 
 	spin_lock(&commit_transaction->t_handle_lock);
-	while (commit_transaction->t_updates) {
+	while (atomic_read(&commit_transaction->t_updates)) {
 		DEFINE_WAIT(wait);
 
 		prepare_to_wait(&journal->j_wait_updates, &wait,
 					TASK_UNINTERRUPTIBLE);
-		if (commit_transaction->t_updates) {
+		if (atomic_read(&commit_transaction->t_updates)) {
 			spin_unlock(&commit_transaction->t_handle_lock);
-			spin_unlock(&journal->j_state_lock);
+			write_unlock(&journal->j_state_lock);
 			schedule();
-			spin_lock(&journal->j_state_lock);
+			write_lock(&journal->j_state_lock);
 			spin_lock(&commit_transaction->t_handle_lock);
 		}
 		finish_wait(&journal->j_wait_updates, &wait);
 	}
 	spin_unlock(&commit_transaction->t_handle_lock);
 
-	J_ASSERT (commit_transaction->t_outstanding_credits <=
+	J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
 			journal->j_max_transaction_buffers);
 
 	/*
@@ -497,7 +497,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 	start_time = ktime_get();
 	commit_transaction->t_log_start = journal->j_head;
 	wake_up(&journal->j_wait_transaction_locked);
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 
 	jbd_debug (3, "JBD: commit phase 2\n");
 
@@ -519,19 +519,20 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 	 * transaction!  Now comes the tricky part: we need to write out
 	 * metadata.  Loop over the transaction's entire buffer list:
 	 */
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_COMMIT;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 
 	trace_jbd2_commit_logging(journal, commit_transaction);
 	stats.run.rs_logging = jiffies;
 	stats.run.rs_flushing = jbd2_time_diff(stats.run.rs_flushing,
 					       stats.run.rs_logging);
-	stats.run.rs_blocks = commit_transaction->t_outstanding_credits;
+	stats.run.rs_blocks =
+		atomic_read(&commit_transaction->t_outstanding_credits);
 	stats.run.rs_blocks_logged = 0;
 
 	J_ASSERT(commit_transaction->t_nr_buffers <=
-		 commit_transaction->t_outstanding_credits);
+		 atomic_read(&commit_transaction->t_outstanding_credits));
 
 	err = 0;
 	descriptor = NULL;
@@ -616,7 +617,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 		 * the free space in the log, but this counter is changed
 		 * by jbd2_journal_next_log_block() also.
 		 */
-		commit_transaction->t_outstanding_credits--;
+		atomic_dec(&commit_transaction->t_outstanding_credits);
 
 		/* Bump b_count to prevent truncate from stumbling over
                    the shadowed buffer!  @@@ This can go if we ever get
@@ -977,7 +978,7 @@ restart_loop:
 	 * __jbd2_journal_drop_transaction(). Otherwise we could race with
 	 * other checkpointing code processing the transaction...
 	 */
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	spin_lock(&journal->j_list_lock);
 	/*
 	 * Now recheck if some buffers did not get attached to the transaction
@@ -985,7 +986,7 @@ restart_loop:
 	 */
 	if (commit_transaction->t_forget) {
 		spin_unlock(&journal->j_list_lock);
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		goto restart_loop;
 	}
 
@@ -1003,7 +1004,8 @@ restart_loop:
 	 * File the transaction statistics
 	 */
 	stats.ts_tid = commit_transaction->t_tid;
-	stats.run.rs_handle_count = commit_transaction->t_handle_count;
+	stats.run.rs_handle_count =
+		atomic_read(&commit_transaction->t_handle_count);
 	trace_jbd2_run_stats(journal->j_fs_dev->bd_dev,
 			     commit_transaction->t_tid, &stats.run);
 
@@ -1037,7 +1039,7 @@ restart_loop:
 				journal->j_average_commit_time*3) / 4;
 	else
 		journal->j_average_commit_time = commit_time;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 
 	if (commit_transaction->t_checkpoint_list == NULL &&
 	    commit_transaction->t_checkpoint_io_list == NULL) {
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 036880895bf..ad5866aaf0f 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -41,6 +41,7 @@
 #include <linux/hash.h>
 #include <linux/log2.h>
 #include <linux/vmalloc.h>
+#include <linux/backing-dev.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/jbd2.h>
@@ -48,8 +49,6 @@
 #include <asm/uaccess.h>
 #include <asm/page.h>
 
-EXPORT_SYMBOL(jbd2_journal_start);
-EXPORT_SYMBOL(jbd2_journal_restart);
 EXPORT_SYMBOL(jbd2_journal_extend);
 EXPORT_SYMBOL(jbd2_journal_stop);
 EXPORT_SYMBOL(jbd2_journal_lock_updates);
@@ -143,7 +142,7 @@ static int kjournald2(void *arg)
 	/*
 	 * And now, wait forever for commit wakeup events.
 	 */
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 
 loop:
 	if (journal->j_flags & JBD2_UNMOUNT)
@@ -154,10 +153,10 @@ loop:
 
 	if (journal->j_commit_sequence != journal->j_commit_request) {
 		jbd_debug(1, "OK, requests differ\n");
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		del_timer_sync(&journal->j_commit_timer);
 		jbd2_journal_commit_transaction(journal);
-		spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
 		goto loop;
 	}
 
@@ -169,9 +168,9 @@ loop:
 		 * be already stopped.
 		 */
 		jbd_debug(1, "Now suspending kjournald2\n");
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		refrigerator();
-		spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
 	} else {
 		/*
 		 * We assume on resume that commits are already there,
@@ -191,9 +190,9 @@ loop:
 		if (journal->j_flags & JBD2_UNMOUNT)
 			should_sleep = 0;
 		if (should_sleep) {
-			spin_unlock(&journal->j_state_lock);
+			write_unlock(&journal->j_state_lock);
 			schedule();
-			spin_lock(&journal->j_state_lock);
+			write_lock(&journal->j_state_lock);
 		}
 		finish_wait(&journal->j_wait_commit, &wait);
 	}
@@ -211,7 +210,7 @@ loop:
 	goto loop;
 
 end_loop:
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	del_timer_sync(&journal->j_commit_timer);
 	journal->j_task = NULL;
 	wake_up(&journal->j_wait_done_commit);
@@ -234,16 +233,16 @@ static int jbd2_journal_start_thread(journal_t *journal)
 
 static void journal_kill_thread(journal_t *journal)
 {
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	journal->j_flags |= JBD2_UNMOUNT;
 
 	while (journal->j_task) {
 		wake_up(&journal->j_wait_commit);
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		wait_event(journal->j_wait_done_commit, journal->j_task == NULL);
-		spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
 	}
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 /*
@@ -310,7 +309,17 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction,
 	 */
 	J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in));
 
-	new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL);
+retry_alloc:
+	new_bh = alloc_buffer_head(GFP_NOFS);
+	if (!new_bh) {
+		/*
+		 * Failure is not an option, but __GFP_NOFAIL is going
+		 * away; so we retry ourselves here.
+		 */
+		congestion_wait(BLK_RW_ASYNC, HZ/50);
+		goto retry_alloc;
+	}
+
 	/* keep subsequent assertions sane */
 	new_bh->b_state = 0;
 	init_buffer(new_bh, NULL, NULL);
@@ -442,7 +451,7 @@ int __jbd2_log_space_left(journal_t *journal)
 {
 	int left = journal->j_free;
 
-	assert_spin_locked(&journal->j_state_lock);
+	/* assert_spin_locked(&journal->j_state_lock); */
 
 	/*
 	 * Be pessimistic here about the number of those free blocks which
@@ -487,9 +496,9 @@ int jbd2_log_start_commit(journal_t *journal, tid_t tid)
 {
 	int ret;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	ret = __jbd2_log_start_commit(journal, tid);
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	return ret;
 }
 
@@ -508,7 +517,7 @@ int jbd2_journal_force_commit_nested(journal_t *journal)
 	transaction_t *transaction = NULL;
 	tid_t tid;
 
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 	if (journal->j_running_transaction && !current->journal_info) {
 		transaction = journal->j_running_transaction;
 		__jbd2_log_start_commit(journal, transaction->t_tid);
@@ -516,12 +525,12 @@ int jbd2_journal_force_commit_nested(journal_t *journal)
 		transaction = journal->j_committing_transaction;
 
 	if (!transaction) {
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		return 0;	/* Nothing to retry */
 	}
 
 	tid = transaction->t_tid;
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 	jbd2_log_wait_commit(journal, tid);
 	return 1;
 }
@@ -535,7 +544,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
 {
 	int ret = 0;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	if (journal->j_running_transaction) {
 		tid_t tid = journal->j_running_transaction->t_tid;
 
@@ -554,7 +563,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
 			*ptid = journal->j_committing_transaction->t_tid;
 		ret = 1;
 	}
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	return ret;
 }
 
@@ -566,26 +575,24 @@ int jbd2_log_wait_commit(journal_t *journal, tid_t tid)
 {
 	int err = 0;
 
+	read_lock(&journal->j_state_lock);
 #ifdef CONFIG_JBD2_DEBUG
-	spin_lock(&journal->j_state_lock);
 	if (!tid_geq(journal->j_commit_request, tid)) {
 		printk(KERN_EMERG
 		       "%s: error: j_commit_request=%d, tid=%d\n",
 		       __func__, journal->j_commit_request, tid);
 	}
-	spin_unlock(&journal->j_state_lock);
 #endif
-	spin_lock(&journal->j_state_lock);
 	while (tid_gt(tid, journal->j_commit_sequence)) {
 		jbd_debug(1, "JBD: want %d, j_commit_sequence=%d\n",
 				  tid, journal->j_commit_sequence);
 		wake_up(&journal->j_wait_commit);
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		wait_event(journal->j_wait_done_commit,
 				!tid_gt(tid, journal->j_commit_sequence));
-		spin_lock(&journal->j_state_lock);
+		read_lock(&journal->j_state_lock);
 	}
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 
 	if (unlikely(is_journal_aborted(journal))) {
 		printk(KERN_EMERG "journal commit I/O error\n");
@@ -602,7 +609,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
 {
 	unsigned long blocknr;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	J_ASSERT(journal->j_free > 1);
 
 	blocknr = journal->j_head;
@@ -610,7 +617,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
 	journal->j_free--;
 	if (journal->j_head == journal->j_last)
 		journal->j_head = journal->j_first;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	return jbd2_journal_bmap(journal, blocknr, retp);
 }
 
@@ -830,7 +837,7 @@ static journal_t * journal_init_common (void)
 	mutex_init(&journal->j_checkpoint_mutex);
 	spin_lock_init(&journal->j_revoke_lock);
 	spin_lock_init(&journal->j_list_lock);
-	spin_lock_init(&journal->j_state_lock);
+	rwlock_init(&journal->j_state_lock);
 
 	journal->j_commit_interval = (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE);
 	journal->j_min_batch_time = 0;
@@ -1096,14 +1103,14 @@ void jbd2_journal_update_superblock(journal_t *journal, int wait)
 		set_buffer_uptodate(bh);
 	}
 
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 	jbd_debug(1,"JBD: updating superblock (start %ld, seq %d, errno %d)\n",
 		  journal->j_tail, journal->j_tail_sequence, journal->j_errno);
 
 	sb->s_sequence = cpu_to_be32(journal->j_tail_sequence);
 	sb->s_start    = cpu_to_be32(journal->j_tail);
 	sb->s_errno    = cpu_to_be32(journal->j_errno);
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 
 	BUFFER_TRACE(bh, "marking dirty");
 	mark_buffer_dirty(bh);
@@ -1124,12 +1131,12 @@ out:
 	 * any future commit will have to be careful to update the
 	 * superblock again to re-record the true start of the log. */
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	if (sb->s_start)
 		journal->j_flags &= ~JBD2_FLUSHED;
 	else
 		journal->j_flags |= JBD2_FLUSHED;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 /*
@@ -1391,13 +1398,9 @@ int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat,
 int jbd2_journal_check_available_features (journal_t *journal, unsigned long compat,
 				      unsigned long ro, unsigned long incompat)
 {
-	journal_superblock_t *sb;
-
 	if (!compat && !ro && !incompat)
 		return 1;
 
-	sb = journal->j_superblock;
-
 	/* We can support any known requested features iff the
 	 * superblock is in version 2.  Otherwise we fail to support any
 	 * extended sb features. */
@@ -1545,7 +1548,7 @@ int jbd2_journal_flush(journal_t *journal)
 	transaction_t *transaction = NULL;
 	unsigned long old_tail;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 
 	/* Force everything buffered to the log... */
 	if (journal->j_running_transaction) {
@@ -1558,10 +1561,10 @@ int jbd2_journal_flush(journal_t *journal)
 	if (transaction) {
 		tid_t tid = transaction->t_tid;
 
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		jbd2_log_wait_commit(journal, tid);
 	} else {
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 	}
 
 	/* ...and flush everything in the log out to disk. */
@@ -1585,12 +1588,12 @@ int jbd2_journal_flush(journal_t *journal)
 	 * the magic code for a fully-recovered superblock.  Any future
 	 * commits of data to the journal will restore the current
 	 * s_start value. */
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	old_tail = journal->j_tail;
 	journal->j_tail = 0;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	jbd2_journal_update_superblock(journal, 1);
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	journal->j_tail = old_tail;
 
 	J_ASSERT(!journal->j_running_transaction);
@@ -1598,7 +1601,7 @@ int jbd2_journal_flush(journal_t *journal)
 	J_ASSERT(!journal->j_checkpoint_transactions);
 	J_ASSERT(journal->j_head == journal->j_tail);
 	J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence);
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	return 0;
 }
 
@@ -1617,7 +1620,6 @@ int jbd2_journal_flush(journal_t *journal)
 
 int jbd2_journal_wipe(journal_t *journal, int write)
 {
-	journal_superblock_t *sb;
 	int err = 0;
 
 	J_ASSERT (!(journal->j_flags & JBD2_LOADED));
@@ -1626,8 +1628,6 @@ int jbd2_journal_wipe(journal_t *journal, int write)
 	if (err)
 		return err;
 
-	sb = journal->j_superblock;
-
 	if (!journal->j_tail)
 		goto no_recovery;
 
@@ -1665,12 +1665,12 @@ void __jbd2_journal_abort_hard(journal_t *journal)
 	printk(KERN_ERR "Aborting journal on device %s.\n",
 	       journal->j_devname);
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	journal->j_flags |= JBD2_ABORT;
 	transaction = journal->j_running_transaction;
 	if (transaction)
 		__jbd2_log_start_commit(journal, transaction->t_tid);
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 /* Soft abort: record the abort error status in the journal superblock,
@@ -1755,12 +1755,12 @@ int jbd2_journal_errno(journal_t *journal)
 {
 	int err;
 
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 	if (journal->j_flags & JBD2_ABORT)
 		err = -EROFS;
 	else
 		err = journal->j_errno;
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 	return err;
 }
 
@@ -1775,12 +1775,12 @@ int jbd2_journal_clear_err(journal_t *journal)
 {
 	int err = 0;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	if (journal->j_flags & JBD2_ABORT)
 		err = -EROFS;
 	else
 		journal->j_errno = 0;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	return err;
 }
 
@@ -1793,10 +1793,10 @@ int jbd2_journal_clear_err(journal_t *journal)
  */
 void jbd2_journal_ack_err(journal_t *journal)
 {
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	if (journal->j_errno)
 		journal->j_flags |= JBD2_ACK_ERR;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 int jbd2_journal_blocks_per_page(struct inode *inode)
@@ -2201,8 +2201,6 @@ void jbd2_journal_init_jbd_inode(struct jbd2_inode *jinode, struct inode *inode)
 void jbd2_journal_release_jbd_inode(journal_t *journal,
 				    struct jbd2_inode *jinode)
 {
-	int writeout = 0;
-
 	if (!journal)
 		return;
 restart:
@@ -2219,9 +2217,6 @@ restart:
 		goto restart;
 	}
 
-	/* Do we need to wait for data writeback? */
-	if (journal->j_committing_transaction == jinode->i_transaction)
-		writeout = 1;
 	if (jinode->i_transaction) {
 		list_del(&jinode->i_list);
 		jinode->i_transaction = NULL;
diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c
index 049281b7cb8..2bc4d5f116f 100644
--- a/fs/jbd2/recovery.c
+++ b/fs/jbd2/recovery.c
@@ -285,12 +285,10 @@ int jbd2_journal_recover(journal_t *journal)
 int jbd2_journal_skip_recovery(journal_t *journal)
 {
 	int			err;
-	journal_superblock_t *	sb;
 
 	struct recovery_info	info;
 
 	memset (&info, 0, sizeof(info));
-	sb = journal->j_superblock;
 
 	err = do_one_pass(journal, &info, PASS_SCAN);
 
@@ -299,7 +297,8 @@ int jbd2_journal_skip_recovery(journal_t *journal)
 		++journal->j_transaction_sequence;
 	} else {
 #ifdef CONFIG_JBD2_DEBUG
-		int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence);
+		int dropped = info.end_transaction - 
+			be32_to_cpu(journal->j_superblock->s_sequence);
 #endif
 		jbd_debug(1,
 			  "JBD: ignoring %d transaction%s from the journal.\n",
@@ -365,11 +364,6 @@ static int do_one_pass(journal_t *journal,
 	int			tag_bytes = journal_tag_bytes(journal);
 	__u32			crc32_sum = ~0; /* Transactional Checksums */
 
-	/* Precompute the maximum metadata descriptors in a descriptor block */
-	int			MAX_BLOCKS_PER_DESC;
-	MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t))
-			       / tag_bytes);
-
 	/*
 	 * First thing is to establish what we expect to find in the log
 	 * (in terms of transaction IDs), and where (in terms of log
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index b8e0806681b..d95cc9d0401 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -26,6 +26,8 @@
 #include <linux/mm.h>
 #include <linux/highmem.h>
 #include <linux/hrtimer.h>
+#include <linux/backing-dev.h>
+#include <linux/module.h>
 
 static void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh);
 
@@ -53,6 +55,9 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction)
 	transaction->t_tid = journal->j_transaction_sequence++;
 	transaction->t_expires = jiffies + journal->j_commit_interval;
 	spin_lock_init(&transaction->t_handle_lock);
+	atomic_set(&transaction->t_updates, 0);
+	atomic_set(&transaction->t_outstanding_credits, 0);
+	atomic_set(&transaction->t_handle_count, 0);
 	INIT_LIST_HEAD(&transaction->t_inode_list);
 	INIT_LIST_HEAD(&transaction->t_private_list);
 
@@ -83,65 +88,75 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction)
  * transaction's buffer credits.
  */
 
-static int start_this_handle(journal_t *journal, handle_t *handle)
+static int start_this_handle(journal_t *journal, handle_t *handle,
+			     int gfp_mask)
 {
 	transaction_t *transaction;
 	int needed;
 	int nblocks = handle->h_buffer_credits;
 	transaction_t *new_transaction = NULL;
-	int ret = 0;
 	unsigned long ts = jiffies;
 
 	if (nblocks > journal->j_max_transaction_buffers) {
 		printk(KERN_ERR "JBD: %s wants too many credits (%d > %d)\n",
 		       current->comm, nblocks,
 		       journal->j_max_transaction_buffers);
-		ret = -ENOSPC;
-		goto out;
+		return -ENOSPC;
 	}
 
 alloc_transaction:
 	if (!journal->j_running_transaction) {
-		new_transaction = kzalloc(sizeof(*new_transaction),
-						GFP_NOFS|__GFP_NOFAIL);
+		new_transaction = kzalloc(sizeof(*new_transaction), gfp_mask);
 		if (!new_transaction) {
-			ret = -ENOMEM;
-			goto out;
+			/*
+			 * If __GFP_FS is not present, then we may be
+			 * being called from inside the fs writeback
+			 * layer, so we MUST NOT fail.  Since
+			 * __GFP_NOFAIL is going away, we will arrange
+			 * to retry the allocation ourselves.
+			 */
+			if ((gfp_mask & __GFP_FS) == 0) {
+				congestion_wait(BLK_RW_ASYNC, HZ/50);
+				goto alloc_transaction;
+			}
+			return -ENOMEM;
 		}
 	}
 
 	jbd_debug(3, "New handle %p going live.\n", handle);
 
-repeat:
-
 	/*
 	 * We need to hold j_state_lock until t_updates has been incremented,
 	 * for proper journal barrier handling
 	 */
-	spin_lock(&journal->j_state_lock);
-repeat_locked:
+repeat:
+	read_lock(&journal->j_state_lock);
 	if (is_journal_aborted(journal) ||
 	    (journal->j_errno != 0 && !(journal->j_flags & JBD2_ACK_ERR))) {
-		spin_unlock(&journal->j_state_lock);
-		ret = -EROFS;
-		goto out;
+		read_unlock(&journal->j_state_lock);
+		kfree(new_transaction);
+		return -EROFS;
 	}
 
 	/* Wait on the journal's transaction barrier if necessary */
 	if (journal->j_barrier_count) {
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		wait_event(journal->j_wait_transaction_locked,
 				journal->j_barrier_count == 0);
 		goto repeat;
 	}
 
 	if (!journal->j_running_transaction) {
-		if (!new_transaction) {
-			spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
+		if (!new_transaction)
 			goto alloc_transaction;
+		write_lock(&journal->j_state_lock);
+		if (!journal->j_running_transaction) {
+			jbd2_get_transaction(journal, new_transaction);
+			new_transaction = NULL;
 		}
-		jbd2_get_transaction(journal, new_transaction);
-		new_transaction = NULL;
+		write_unlock(&journal->j_state_lock);
+		goto repeat;
 	}
 
 	transaction = journal->j_running_transaction;
@@ -155,7 +170,7 @@ repeat_locked:
 
 		prepare_to_wait(&journal->j_wait_transaction_locked,
 					&wait, TASK_UNINTERRUPTIBLE);
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		schedule();
 		finish_wait(&journal->j_wait_transaction_locked, &wait);
 		goto repeat;
@@ -166,8 +181,8 @@ repeat_locked:
 	 * buffers requested by this operation, we need to stall pending a log
 	 * checkpoint to free some more log space.
 	 */
-	spin_lock(&transaction->t_handle_lock);
-	needed = transaction->t_outstanding_credits + nblocks;
+	needed = atomic_add_return(nblocks,
+				   &transaction->t_outstanding_credits);
 
 	if (needed > journal->j_max_transaction_buffers) {
 		/*
@@ -178,11 +193,11 @@ repeat_locked:
 		DEFINE_WAIT(wait);
 
 		jbd_debug(2, "Handle %p starting new commit...\n", handle);
-		spin_unlock(&transaction->t_handle_lock);
+		atomic_sub(nblocks, &transaction->t_outstanding_credits);
 		prepare_to_wait(&journal->j_wait_transaction_locked, &wait,
 				TASK_UNINTERRUPTIBLE);
 		__jbd2_log_start_commit(journal, transaction->t_tid);
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 		schedule();
 		finish_wait(&journal->j_wait_transaction_locked, &wait);
 		goto repeat;
@@ -215,35 +230,48 @@ repeat_locked:
 	 */
 	if (__jbd2_log_space_left(journal) < jbd_space_needed(journal)) {
 		jbd_debug(2, "Handle %p waiting for checkpoint...\n", handle);
-		spin_unlock(&transaction->t_handle_lock);
-		__jbd2_log_wait_for_space(journal);
-		goto repeat_locked;
+		atomic_sub(nblocks, &transaction->t_outstanding_credits);
+		read_unlock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
+		if (__jbd2_log_space_left(journal) < jbd_space_needed(journal))
+			__jbd2_log_wait_for_space(journal);
+		write_unlock(&journal->j_state_lock);
+		goto repeat;
 	}
 
 	/* OK, account for the buffers that this operation expects to
-	 * use and add the handle to the running transaction. */
-
-	if (time_after(transaction->t_start, ts)) {
+	 * use and add the handle to the running transaction. 
+	 *
+	 * In order for t_max_wait to be reliable, it must be
+	 * protected by a lock.  But doing so will mean that
+	 * start_this_handle() can not be run in parallel on SMP
+	 * systems, which limits our scalability.  So we only enable
+	 * it when debugging is enabled.  We may want to use a
+	 * separate flag, eventually, so we can enable this
+	 * independently of debugging.
+	 */
+#ifdef CONFIG_JBD2_DEBUG
+	if (jbd2_journal_enable_debug &&
+	    time_after(transaction->t_start, ts)) {
 		ts = jbd2_time_diff(ts, transaction->t_start);
+		spin_lock(&transaction->t_handle_lock);
 		if (ts > transaction->t_max_wait)
 			transaction->t_max_wait = ts;
+		spin_unlock(&transaction->t_handle_lock);
 	}
-
+#endif
 	handle->h_transaction = transaction;
-	transaction->t_outstanding_credits += nblocks;
-	transaction->t_updates++;
-	transaction->t_handle_count++;
+	atomic_inc(&transaction->t_updates);
+	atomic_inc(&transaction->t_handle_count);
 	jbd_debug(4, "Handle %p given %d credits (total %d, free %d)\n",
-		  handle, nblocks, transaction->t_outstanding_credits,
+		  handle, nblocks,
+		  atomic_read(&transaction->t_outstanding_credits),
 		  __jbd2_log_space_left(journal));
-	spin_unlock(&transaction->t_handle_lock);
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 
 	lock_map_acquire(&handle->h_lockdep_map);
-out:
-	if (unlikely(new_transaction))		/* It's usually NULL */
-		kfree(new_transaction);
-	return ret;
+	kfree(new_transaction);
+	return 0;
 }
 
 static struct lock_class_key jbd2_handle_key;
@@ -278,7 +306,7 @@ static handle_t *new_handle(int nblocks)
  *
  * Return a pointer to a newly allocated handle, or NULL on failure
  */
-handle_t *jbd2_journal_start(journal_t *journal, int nblocks)
+handle_t *jbd2__journal_start(journal_t *journal, int nblocks, int gfp_mask)
 {
 	handle_t *handle = journal_current_handle();
 	int err;
@@ -298,7 +326,7 @@ handle_t *jbd2_journal_start(journal_t *journal, int nblocks)
 
 	current->journal_info = handle;
 
-	err = start_this_handle(journal, handle);
+	err = start_this_handle(journal, handle, gfp_mask);
 	if (err < 0) {
 		jbd2_free_handle(handle);
 		current->journal_info = NULL;
@@ -308,6 +336,15 @@ handle_t *jbd2_journal_start(journal_t *journal, int nblocks)
 out:
 	return handle;
 }
+EXPORT_SYMBOL(jbd2__journal_start);
+
+
+handle_t *jbd2_journal_start(journal_t *journal, int nblocks)
+{
+	return jbd2__journal_start(journal, nblocks, GFP_NOFS);
+}
+EXPORT_SYMBOL(jbd2_journal_start);
+
 
 /**
  * int jbd2_journal_extend() - extend buffer credits.
@@ -342,7 +379,7 @@ int jbd2_journal_extend(handle_t *handle, int nblocks)
 
 	result = 1;
 
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 
 	/* Don't extend a locked-down transaction! */
 	if (handle->h_transaction->t_state != T_RUNNING) {
@@ -352,7 +389,7 @@ int jbd2_journal_extend(handle_t *handle, int nblocks)
 	}
 
 	spin_lock(&transaction->t_handle_lock);
-	wanted = transaction->t_outstanding_credits + nblocks;
+	wanted = atomic_read(&transaction->t_outstanding_credits) + nblocks;
 
 	if (wanted > journal->j_max_transaction_buffers) {
 		jbd_debug(3, "denied handle %p %d blocks: "
@@ -367,14 +404,14 @@ int jbd2_journal_extend(handle_t *handle, int nblocks)
 	}
 
 	handle->h_buffer_credits += nblocks;
-	transaction->t_outstanding_credits += nblocks;
+	atomic_add(nblocks, &transaction->t_outstanding_credits);
 	result = 0;
 
 	jbd_debug(3, "extended handle %p by %d\n", handle, nblocks);
 unlock:
 	spin_unlock(&transaction->t_handle_lock);
 error_out:
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 out:
 	return result;
 }
@@ -394,8 +431,7 @@ out:
  * transaction capabable of guaranteeing the requested number of
  * credits.
  */
-
-int jbd2_journal_restart(handle_t *handle, int nblocks)
+int jbd2__journal_restart(handle_t *handle, int nblocks, int gfp_mask)
 {
 	transaction_t *transaction = handle->h_transaction;
 	journal_t *journal = transaction->t_journal;
@@ -410,29 +446,35 @@ int jbd2_journal_restart(handle_t *handle, int nblocks)
 	 * First unlink the handle from its current transaction, and start the
 	 * commit on that.
 	 */
-	J_ASSERT(transaction->t_updates > 0);
+	J_ASSERT(atomic_read(&transaction->t_updates) > 0);
 	J_ASSERT(journal_current_handle() == handle);
 
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 	spin_lock(&transaction->t_handle_lock);
-	transaction->t_outstanding_credits -= handle->h_buffer_credits;
-	transaction->t_updates--;
-
-	if (!transaction->t_updates)
+	atomic_sub(handle->h_buffer_credits,
+		   &transaction->t_outstanding_credits);
+	if (atomic_dec_and_test(&transaction->t_updates))
 		wake_up(&journal->j_wait_updates);
 	spin_unlock(&transaction->t_handle_lock);
 
 	jbd_debug(2, "restarting handle %p\n", handle);
 	__jbd2_log_start_commit(journal, transaction->t_tid);
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 
 	lock_map_release(&handle->h_lockdep_map);
 	handle->h_buffer_credits = nblocks;
-	ret = start_this_handle(journal, handle);
+	ret = start_this_handle(journal, handle, gfp_mask);
 	return ret;
 }
+EXPORT_SYMBOL(jbd2__journal_restart);
 
 
+int jbd2_journal_restart(handle_t *handle, int nblocks)
+{
+	return jbd2__journal_restart(handle, nblocks, GFP_NOFS);
+}
+EXPORT_SYMBOL(jbd2_journal_restart);
+
 /**
  * void jbd2_journal_lock_updates () - establish a transaction barrier.
  * @journal:  Journal to establish a barrier on.
@@ -447,7 +489,7 @@ void jbd2_journal_lock_updates(journal_t *journal)
 {
 	DEFINE_WAIT(wait);
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	++journal->j_barrier_count;
 
 	/* Wait until there are no running updates */
@@ -458,19 +500,19 @@ void jbd2_journal_lock_updates(journal_t *journal)
 			break;
 
 		spin_lock(&transaction->t_handle_lock);
-		if (!transaction->t_updates) {
+		if (!atomic_read(&transaction->t_updates)) {
 			spin_unlock(&transaction->t_handle_lock);
 			break;
 		}
 		prepare_to_wait(&journal->j_wait_updates, &wait,
 				TASK_UNINTERRUPTIBLE);
 		spin_unlock(&transaction->t_handle_lock);
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		schedule();
 		finish_wait(&journal->j_wait_updates, &wait);
-		spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
 	}
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 
 	/*
 	 * We have now established a barrier against other normal updates, but
@@ -494,9 +536,9 @@ void jbd2_journal_unlock_updates (journal_t *journal)
 	J_ASSERT(journal->j_barrier_count != 0);
 
 	mutex_unlock(&journal->j_barrier);
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	--journal->j_barrier_count;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 	wake_up(&journal->j_wait_transaction_locked);
 }
 
@@ -1238,7 +1280,8 @@ int jbd2_journal_stop(handle_t *handle)
 {
 	transaction_t *transaction = handle->h_transaction;
 	journal_t *journal = transaction->t_journal;
-	int err;
+	int err, wait_for_commit = 0;
+	tid_t tid;
 	pid_t pid;
 
 	J_ASSERT(journal_current_handle() == handle);
@@ -1246,7 +1289,7 @@ int jbd2_journal_stop(handle_t *handle)
 	if (is_handle_aborted(handle))
 		err = -EIO;
 	else {
-		J_ASSERT(transaction->t_updates > 0);
+		J_ASSERT(atomic_read(&transaction->t_updates) > 0);
 		err = 0;
 	}
 
@@ -1291,9 +1334,9 @@ int jbd2_journal_stop(handle_t *handle)
 
 		journal->j_last_sync_writer = pid;
 
-		spin_lock(&journal->j_state_lock);
+		read_lock(&journal->j_state_lock);
 		commit_time = journal->j_average_commit_time;
-		spin_unlock(&journal->j_state_lock);
+		read_unlock(&journal->j_state_lock);
 
 		trans_time = ktime_to_ns(ktime_sub(ktime_get(),
 						   transaction->t_start_time));
@@ -1314,14 +1357,8 @@ int jbd2_journal_stop(handle_t *handle)
 	if (handle->h_sync)
 		transaction->t_synchronous_commit = 1;
 	current->journal_info = NULL;
-	spin_lock(&transaction->t_handle_lock);
-	transaction->t_outstanding_credits -= handle->h_buffer_credits;
-	transaction->t_updates--;
-	if (!transaction->t_updates) {
-		wake_up(&journal->j_wait_updates);
-		if (journal->j_barrier_count)
-			wake_up(&journal->j_wait_transaction_locked);
-	}
+	atomic_sub(handle->h_buffer_credits,
+		   &transaction->t_outstanding_credits);
 
 	/*
 	 * If the handle is marked SYNC, we need to set another commit
@@ -1330,15 +1367,13 @@ int jbd2_journal_stop(handle_t *handle)
 	 * transaction is too old now.
 	 */
 	if (handle->h_sync ||
-			transaction->t_outstanding_credits >
-				journal->j_max_transaction_buffers ||
-			time_after_eq(jiffies, transaction->t_expires)) {
+	    (atomic_read(&transaction->t_outstanding_credits) >
+	     journal->j_max_transaction_buffers) ||
+	    time_after_eq(jiffies, transaction->t_expires)) {
 		/* Do this even for aborted journals: an abort still
 		 * completes the commit thread, it just doesn't write
 		 * anything to disk. */
-		tid_t tid = transaction->t_tid;
 
-		spin_unlock(&transaction->t_handle_lock);
 		jbd_debug(2, "transaction too old, requesting commit for "
 					"handle %p\n", handle);
 		/* This is non-blocking */
@@ -1349,11 +1384,25 @@ int jbd2_journal_stop(handle_t *handle)
 		 * to wait for the commit to complete.
 		 */
 		if (handle->h_sync && !(current->flags & PF_MEMALLOC))
-			err = jbd2_log_wait_commit(journal, tid);
-	} else {
-		spin_unlock(&transaction->t_handle_lock);
+			wait_for_commit = 1;
 	}
 
+	/*
+	 * Once we drop t_updates, if it goes to zero the transaction
+	 * could start commiting on us and eventually disappear.  So
+	 * once we do this, we must not dereference transaction
+	 * pointer again.
+	 */
+	tid = transaction->t_tid;
+	if (atomic_dec_and_test(&transaction->t_updates)) {
+		wake_up(&journal->j_wait_updates);
+		if (journal->j_barrier_count)
+			wake_up(&journal->j_wait_transaction_locked);
+	}
+
+	if (wait_for_commit)
+		err = jbd2_log_wait_commit(journal, tid);
+
 	lock_map_release(&handle->h_lockdep_map);
 
 	jbd2_free_handle(handle);
@@ -1719,7 +1768,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 		goto zap_buffer_unlocked;
 
 	/* OK, we have data buffer in journaled mode */
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	jbd_lock_bh_state(bh);
 	spin_lock(&journal->j_list_lock);
 
@@ -1772,7 +1821,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 			jbd2_journal_put_journal_head(jh);
 			spin_unlock(&journal->j_list_lock);
 			jbd_unlock_bh_state(bh);
-			spin_unlock(&journal->j_state_lock);
+			write_unlock(&journal->j_state_lock);
 			return ret;
 		} else {
 			/* There is no currently-running transaction. So the
@@ -1786,7 +1835,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 				jbd2_journal_put_journal_head(jh);
 				spin_unlock(&journal->j_list_lock);
 				jbd_unlock_bh_state(bh);
-				spin_unlock(&journal->j_state_lock);
+				write_unlock(&journal->j_state_lock);
 				return ret;
 			} else {
 				/* The orphan record's transaction has
@@ -1810,7 +1859,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 		jbd2_journal_put_journal_head(jh);
 		spin_unlock(&journal->j_list_lock);
 		jbd_unlock_bh_state(bh);
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		return 0;
 	} else {
 		/* Good, the buffer belongs to the running transaction.
@@ -1829,7 +1878,7 @@ zap_buffer:
 zap_buffer_no_jh:
 	spin_unlock(&journal->j_list_lock);
 	jbd_unlock_bh_state(bh);
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 zap_buffer_unlocked:
 	clear_buffer_dirty(bh);
 	J_ASSERT_BH(bh, !buffer_jbddirty(bh));
@@ -2136,9 +2185,9 @@ int jbd2_journal_begin_ordered_truncate(journal_t *journal,
 	/* Locks are here just to force reading of recent values, it is
 	 * enough that the transaction was not committing before we started
 	 * a transaction adding the inode to orphan list */
-	spin_lock(&journal->j_state_lock);
+	read_lock(&journal->j_state_lock);
 	commit_trans = journal->j_committing_transaction;
-	spin_unlock(&journal->j_state_lock);
+	read_unlock(&journal->j_state_lock);
 	spin_lock(&journal->j_list_lock);
 	inode_trans = jinode->i_transaction;
 	spin_unlock(&journal->j_list_lock);
diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig
index a43d07e7b92..cc1bb33b59b 100644
--- a/fs/nfs/Kconfig
+++ b/fs/nfs/Kconfig
@@ -61,8 +61,8 @@ config NFS_V3_ACL
 	  If unsure, say N.
 
 config NFS_V4
-	bool "NFS client support for NFS version 4 (EXPERIMENTAL)"
-	depends on NFS_FS && EXPERIMENTAL
+	bool "NFS client support for NFS version 4"
+	depends on NFS_FS
 	select RPCSEC_GSS_KRB5
 	help
 	  This option enables support for version 4 of the NFS protocol
@@ -72,16 +72,16 @@ config NFS_V4
 	  space programs which can be found in the Linux nfs-utils package,
 	  available from http://linux-nfs.org/.
 
-	  If unsure, say N.
+	  If unsure, say Y.
 
 config NFS_V4_1
-	bool "NFS client support for NFSv4.1 (DEVELOPER ONLY)"
+	bool "NFS client support for NFSv4.1 (EXPERIMENTAL)"
 	depends on NFS_V4 && EXPERIMENTAL
 	help
 	  This option enables support for minor version 1 of the NFSv4 protocol
 	  (draft-ietf-nfsv4-minorversion1) in the kernel's NFS client.
 
-	  Unless you're an NFS developer, say N.
+	  If unsure, say N.
 
 config ROOT_NFS
 	bool "Root file system on NFS"
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index a08770a7e85..930d10fecda 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -37,8 +37,8 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *
 	if (inode == NULL)
 		goto out_putclient;
 	nfsi = NFS_I(inode);
-	down_read(&nfsi->rwsem);
-	delegation = nfsi->delegation;
+	rcu_read_lock();
+	delegation = rcu_dereference(nfsi->delegation);
 	if (delegation == NULL || (delegation->type & FMODE_WRITE) == 0)
 		goto out_iput;
 	res->size = i_size_read(inode);
@@ -53,7 +53,7 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *
 		args->bitmap[1];
 	res->status = 0;
 out_iput:
-	up_read(&nfsi->rwsem);
+	rcu_read_unlock();
 	iput(inode);
 out_putclient:
 	nfs_put_client(clp);
@@ -62,16 +62,6 @@ out:
 	return res->status;
 }
 
-static int (*nfs_validate_delegation_stateid(struct nfs_client *clp))(struct nfs_delegation *, const nfs4_stateid *)
-{
-#if defined(CONFIG_NFS_V4_1)
-	if (clp->cl_minorversion > 0)
-		return nfs41_validate_delegation_stateid;
-#endif
-	return nfs4_validate_delegation_stateid;
-}
-
-
 __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 {
 	struct nfs_client *clp;
@@ -92,8 +82,7 @@ __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 		inode = nfs_delegation_find_inode(clp, &args->fh);
 		if (inode != NULL) {
 			/* Set up a helper thread to actually return the delegation */
-			switch (nfs_async_inode_return_delegation(inode, &args->stateid,
-								  nfs_validate_delegation_stateid(clp))) {
+			switch (nfs_async_inode_return_delegation(inode, &args->stateid)) {
 				case 0:
 					res = 0;
 					break;
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index d25b5257b7a..4e7df2adb21 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -150,6 +150,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_
 	clp->cl_boot_time = CURRENT_TIME;
 	clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;
 	clp->cl_minorversion = cl_init->minorversion;
+	clp->cl_mvops = nfs_v4_minor_ops[cl_init->minorversion];
 #endif
 	cred = rpc_lookup_machine_cred();
 	if (!IS_ERR(cred))
@@ -178,7 +179,7 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp)
 		clp->cl_session = NULL;
 	}
 
-	clp->cl_call_sync = _nfs4_call_sync;
+	clp->cl_mvops = nfs_v4_minor_ops[0];
 #endif /* CONFIG_NFS_V4_1 */
 }
 
@@ -188,7 +189,7 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp)
 static void nfs4_destroy_callback(struct nfs_client *clp)
 {
 	if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
-		nfs_callback_down(clp->cl_minorversion);
+		nfs_callback_down(clp->cl_mvops->minor_version);
 }
 
 static void nfs4_shutdown_client(struct nfs_client *clp)
@@ -1126,7 +1127,7 @@ static int nfs4_init_callback(struct nfs_client *clp)
 				return error;
 		}
 
-		error = nfs_callback_up(clp->cl_minorversion,
+		error = nfs_callback_up(clp->cl_mvops->minor_version,
 					clp->cl_rpcclient->cl_xprt);
 		if (error < 0) {
 			dprintk("%s: failed to start callback. Error = %d\n",
@@ -1143,10 +1144,8 @@ static int nfs4_init_callback(struct nfs_client *clp)
  */
 static int nfs4_init_client_minor_version(struct nfs_client *clp)
 {
-	clp->cl_call_sync = _nfs4_call_sync;
-
 #if defined(CONFIG_NFS_V4_1)
-	if (clp->cl_minorversion) {
+	if (clp->cl_mvops->minor_version) {
 		struct nfs4_session *session = NULL;
 		/*
 		 * Create the session and mark it expired.
@@ -1158,7 +1157,13 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp)
 			return -ENOMEM;
 
 		clp->cl_session = session;
-		clp->cl_call_sync = _nfs4_call_sync_session;
+		/*
+		 * The create session reply races with the server back
+		 * channel probe. Mark the client NFS_CS_SESSION_INITING
+		 * so that the client back channel can find the
+		 * nfs_client struct
+		 */
+		clp->cl_cons_state = NFS_CS_SESSION_INITING;
 	}
 #endif /* CONFIG_NFS_V4_1 */
 
@@ -1454,7 +1459,7 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data,
 				data->authflavor,
 				parent_server->client->cl_xprt->prot,
 				parent_server->client->cl_timeout,
-				parent_client->cl_minorversion);
+				parent_client->cl_mvops->minor_version);
 	if (error < 0)
 		goto error;
 
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 30163454397..b9c3c43cea1 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -268,14 +268,6 @@ out:
 	return status;
 }
 
-/* Sync all data to disk upon delegation return */
-static void nfs_msync_inode(struct inode *inode)
-{
-	filemap_fdatawrite(inode->i_mapping);
-	nfs_wb_all(inode);
-	filemap_fdatawait(inode->i_mapping);
-}
-
 /*
  * Basic procedure for returning a delegation to the server
  */
@@ -367,7 +359,7 @@ int nfs_inode_return_delegation(struct inode *inode)
 		delegation = nfs_detach_delegation_locked(nfsi, NULL, clp);
 		spin_unlock(&clp->cl_lock);
 		if (delegation != NULL) {
-			nfs_msync_inode(inode);
+			nfs_wb_all(inode);
 			err = __nfs_inode_return_delegation(inode, delegation, 1);
 		}
 	}
@@ -471,9 +463,7 @@ void nfs_expire_unreferenced_delegations(struct nfs_client *clp)
 /*
  * Asynchronous delegation recall!
  */
-int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid,
-				      int (*validate_stateid)(struct nfs_delegation *delegation,
-							      const nfs4_stateid *stateid))
+int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid)
 {
 	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_delegation *delegation;
@@ -481,7 +471,7 @@ int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *s
 	rcu_read_lock();
 	delegation = rcu_dereference(NFS_I(inode)->delegation);
 
-	if (!validate_stateid(delegation, stateid)) {
+	if (!clp->cl_mvops->validate_stateid(delegation, stateid)) {
 		rcu_read_unlock();
 		return -ENOENT;
 	}
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index 69e7b814012..2026304bda1 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -34,9 +34,7 @@ enum {
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res);
 void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res);
 int nfs_inode_return_delegation(struct inode *inode);
-int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid,
-				      int (*validate_stateid)(struct nfs_delegation *delegation,
-							      const nfs4_stateid *stateid));
+int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid);
 void nfs_inode_return_delegation_noreclaim(struct inode *inode);
 
 struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle);
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 832e9e23932..29539ceeb74 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1652,16 +1652,7 @@ static int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		}
 	}
 
-	/*
-	 * ... prune child dentries and writebacks if needed.
-	 */
-	if (atomic_read(&old_dentry->d_count) > 1) {
-		if (S_ISREG(old_inode->i_mode))
-			nfs_wb_all(old_inode);
-		shrink_dcache_parent(old_dentry);
-	}
 	nfs_inode_return_delegation(old_inode);
-
 	if (new_inode != NULL)
 		nfs_inode_return_delegation(new_inode);
 
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index ad4cd31d605..064a8096167 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -69,6 +69,7 @@ struct nfs_direct_req {
 
 	/* I/O parameters */
 	struct nfs_open_context	*ctx;		/* file open context info */
+	struct nfs_lock_context *l_ctx;		/* Lock context info */
 	struct kiocb *		iocb;		/* controlling i/o request */
 	struct inode *		inode;		/* target file of i/o */
 
@@ -160,6 +161,7 @@ static inline struct nfs_direct_req *nfs_direct_req_alloc(void)
 	INIT_LIST_HEAD(&dreq->rewrite_list);
 	dreq->iocb = NULL;
 	dreq->ctx = NULL;
+	dreq->l_ctx = NULL;
 	spin_lock_init(&dreq->lock);
 	atomic_set(&dreq->io_count, 0);
 	dreq->count = 0;
@@ -173,6 +175,8 @@ static void nfs_direct_req_free(struct kref *kref)
 {
 	struct nfs_direct_req *dreq = container_of(kref, struct nfs_direct_req, kref);
 
+	if (dreq->l_ctx != NULL)
+		nfs_put_lock_context(dreq->l_ctx);
 	if (dreq->ctx != NULL)
 		put_nfs_open_context(dreq->ctx);
 	kmem_cache_free(nfs_direct_cachep, dreq);
@@ -336,6 +340,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq,
 		data->cred = msg.rpc_cred;
 		data->args.fh = NFS_FH(inode);
 		data->args.context = ctx;
+		data->args.lock_context = dreq->l_ctx;
 		data->args.offset = pos;
 		data->args.pgbase = pgbase;
 		data->args.pages = data->pagevec;
@@ -416,24 +421,28 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nfs_direct_req *dreq,
 static ssize_t nfs_direct_read(struct kiocb *iocb, const struct iovec *iov,
 			       unsigned long nr_segs, loff_t pos)
 {
-	ssize_t result = 0;
+	ssize_t result = -ENOMEM;
 	struct inode *inode = iocb->ki_filp->f_mapping->host;
 	struct nfs_direct_req *dreq;
 
 	dreq = nfs_direct_req_alloc();
-	if (!dreq)
-		return -ENOMEM;
+	if (dreq == NULL)
+		goto out;
 
 	dreq->inode = inode;
 	dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp));
+	dreq->l_ctx = nfs_get_lock_context(dreq->ctx);
+	if (dreq->l_ctx == NULL)
+		goto out_release;
 	if (!is_sync_kiocb(iocb))
 		dreq->iocb = iocb;
 
 	result = nfs_direct_read_schedule_iovec(dreq, iov, nr_segs, pos);
 	if (!result)
 		result = nfs_direct_wait(dreq);
+out_release:
 	nfs_direct_req_release(dreq);
-
+out:
 	return result;
 }
 
@@ -574,6 +583,7 @@ static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq)
 	data->args.offset = 0;
 	data->args.count = 0;
 	data->args.context = dreq->ctx;
+	data->args.lock_context = dreq->l_ctx;
 	data->res.count = 0;
 	data->res.fattr = &data->fattr;
 	data->res.verf = &data->verf;
@@ -761,6 +771,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq,
 		data->cred = msg.rpc_cred;
 		data->args.fh = NFS_FH(inode);
 		data->args.context = ctx;
+		data->args.lock_context = dreq->l_ctx;
 		data->args.offset = pos;
 		data->args.pgbase = pgbase;
 		data->args.pages = data->pagevec;
@@ -845,7 +856,7 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov,
 				unsigned long nr_segs, loff_t pos,
 				size_t count)
 {
-	ssize_t result = 0;
+	ssize_t result = -ENOMEM;
 	struct inode *inode = iocb->ki_filp->f_mapping->host;
 	struct nfs_direct_req *dreq;
 	size_t wsize = NFS_SERVER(inode)->wsize;
@@ -853,7 +864,7 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov,
 
 	dreq = nfs_direct_req_alloc();
 	if (!dreq)
-		return -ENOMEM;
+		goto out;
 	nfs_alloc_commit_data(dreq);
 
 	if (dreq->commit_data == NULL || count < wsize)
@@ -861,14 +872,18 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov,
 
 	dreq->inode = inode;
 	dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp));
+	dreq->l_ctx = nfs_get_lock_context(dreq->ctx);
+	if (dreq->l_ctx != NULL)
+		goto out_release;
 	if (!is_sync_kiocb(iocb))
 		dreq->iocb = iocb;
 
 	result = nfs_direct_write_schedule_iovec(dreq, iov, nr_segs, pos, sync);
 	if (!result)
 		result = nfs_direct_wait(dreq);
+out_release:
 	nfs_direct_req_release(dreq);
-
+out:
 	return result;
 }
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index f036153d9f5..2d141a74ae8 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -203,37 +203,11 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 }
 
 /*
- * Helper for nfs_file_flush() and nfs_file_fsync()
- *
- * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to
- * disk, but it retrieves and clears ctx->error after synching, despite
- * the two being set at the same time in nfs_context_set_write_error().
- * This is because the former is used to notify the _next_ call to
- * nfs_file_write() that a write error occured, and hence cause it to
- * fall back to doing a synchronous write.
- */
-static int nfs_do_fsync(struct nfs_open_context *ctx, struct inode *inode)
-{
-	int have_error, status;
-	int ret = 0;
-
-	have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
-	status = nfs_wb_all(inode);
-	have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
-	if (have_error)
-		ret = xchg(&ctx->error, 0);
-	if (!ret)
-		ret = status;
-	return ret;
-}
-
-/*
  * Flush all dirty pages, and check for write errors.
  */
 static int
 nfs_file_flush(struct file *file, fl_owner_t id)
 {
-	struct nfs_open_context *ctx = nfs_file_open_context(file);
 	struct dentry	*dentry = file->f_path.dentry;
 	struct inode	*inode = dentry->d_inode;
 
@@ -246,7 +220,7 @@ nfs_file_flush(struct file *file, fl_owner_t id)
 		return 0;
 
 	/* Flush writes to the server and return any errors */
-	return nfs_do_fsync(ctx, inode);
+	return vfs_fsync(file, 0);
 }
 
 static ssize_t
@@ -321,6 +295,13 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
  * Flush any dirty pages for this process, and check for write errors.
  * The return status from this call provides a reliable indication of
  * whether any write errors occurred for this process.
+ *
+ * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to
+ * disk, but it retrieves and clears ctx->error after synching, despite
+ * the two being set at the same time in nfs_context_set_write_error().
+ * This is because the former is used to notify the _next_ call to
+ * nfs_file_write() that a write error occured, and hence cause it to
+ * fall back to doing a synchronous write.
  */
 static int
 nfs_file_fsync(struct file *file, int datasync)
@@ -328,13 +309,23 @@ nfs_file_fsync(struct file *file, int datasync)
 	struct dentry *dentry = file->f_path.dentry;
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
 	struct inode *inode = dentry->d_inode;
+	int have_error, status;
+	int ret = 0;
+
 
 	dprintk("NFS: fsync file(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
 	nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
-	return nfs_do_fsync(ctx, inode);
+	have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
+	status = nfs_commit_inode(inode, FLUSH_SYNC);
+	have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
+	if (have_error)
+		ret = xchg(&ctx->error, 0);
+	if (!ret)
+		ret = status;
+	return ret;
 }
 
 /*
@@ -648,7 +639,7 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov,
 
 	/* Return error values for O_DSYNC and IS_SYNC() */
 	if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) {
-		int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode);
+		int err = vfs_fsync(iocb->ki_filp, 0);
 		if (err < 0)
 			result = err;
 	}
@@ -684,7 +675,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
 		written = ret;
 
 	if (ret >= 0 && nfs_need_sync_write(filp, inode)) {
-		int err = nfs_do_fsync(nfs_file_open_context(filp), inode);
+		int err = vfs_fsync(filp, 0);
 		if (err < 0)
 			ret = err;
 	}
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 099b3518fee..581d8f081e6 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -413,10 +413,8 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr)
 		return 0;
 
 	/* Write all dirty data */
-	if (S_ISREG(inode->i_mode)) {
-		filemap_write_and_wait(inode->i_mapping);
+	if (S_ISREG(inode->i_mode))
 		nfs_wb_all(inode);
-	}
 
 	fattr = nfs_alloc_fattr();
 	if (fattr == NULL)
@@ -530,6 +528,68 @@ out:
 	return err;
 }
 
+static void nfs_init_lock_context(struct nfs_lock_context *l_ctx)
+{
+	atomic_set(&l_ctx->count, 1);
+	l_ctx->lockowner = current->files;
+	l_ctx->pid = current->tgid;
+	INIT_LIST_HEAD(&l_ctx->list);
+}
+
+static struct nfs_lock_context *__nfs_find_lock_context(struct nfs_open_context *ctx)
+{
+	struct nfs_lock_context *pos;
+
+	list_for_each_entry(pos, &ctx->lock_context.list, list) {
+		if (pos->lockowner != current->files)
+			continue;
+		if (pos->pid != current->tgid)
+			continue;
+		atomic_inc(&pos->count);
+		return pos;
+	}
+	return NULL;
+}
+
+struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx)
+{
+	struct nfs_lock_context *res, *new = NULL;
+	struct inode *inode = ctx->path.dentry->d_inode;
+
+	spin_lock(&inode->i_lock);
+	res = __nfs_find_lock_context(ctx);
+	if (res == NULL) {
+		spin_unlock(&inode->i_lock);
+		new = kmalloc(sizeof(*new), GFP_KERNEL);
+		if (new == NULL)
+			return NULL;
+		nfs_init_lock_context(new);
+		spin_lock(&inode->i_lock);
+		res = __nfs_find_lock_context(ctx);
+		if (res == NULL) {
+			list_add_tail(&new->list, &ctx->lock_context.list);
+			new->open_context = ctx;
+			res = new;
+			new = NULL;
+		}
+	}
+	spin_unlock(&inode->i_lock);
+	kfree(new);
+	return res;
+}
+
+void nfs_put_lock_context(struct nfs_lock_context *l_ctx)
+{
+	struct nfs_open_context *ctx = l_ctx->open_context;
+	struct inode *inode = ctx->path.dentry->d_inode;
+
+	if (!atomic_dec_and_lock(&l_ctx->count, &inode->i_lock))
+		return;
+	list_del(&l_ctx->list);
+	spin_unlock(&inode->i_lock);
+	kfree(l_ctx);
+}
+
 /**
  * nfs_close_context - Common close_context() routine NFSv2/v3
  * @ctx: pointer to context
@@ -566,11 +626,11 @@ static struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct
 		path_get(&ctx->path);
 		ctx->cred = get_rpccred(cred);
 		ctx->state = NULL;
-		ctx->lockowner = current->files;
 		ctx->flags = 0;
 		ctx->error = 0;
 		ctx->dir_cookie = 0;
-		atomic_set(&ctx->count, 1);
+		nfs_init_lock_context(&ctx->lock_context);
+		ctx->lock_context.open_context = ctx;
 	}
 	return ctx;
 }
@@ -578,7 +638,7 @@ static struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct
 struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx)
 {
 	if (ctx != NULL)
-		atomic_inc(&ctx->count);
+		atomic_inc(&ctx->lock_context.count);
 	return ctx;
 }
 
@@ -586,7 +646,7 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 {
 	struct inode *inode = ctx->path.dentry->d_inode;
 
-	if (!atomic_dec_and_lock(&ctx->count, &inode->i_lock))
+	if (!atomic_dec_and_lock(&ctx->lock_context.count, &inode->i_lock))
 		return;
 	list_del(&ctx->list);
 	spin_unlock(&inode->i_lock);
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index e70f44b9b3f..4c2150d8671 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -370,10 +370,9 @@ unsigned int nfs_page_array_len(unsigned int base, size_t len)
  * Helper for restarting RPC calls in the possible presence of NFSv4.1
  * sessions.
  */
-static inline void nfs_restart_rpc(struct rpc_task *task, const struct nfs_client *clp)
+static inline int nfs_restart_rpc(struct rpc_task *task, const struct nfs_client *clp)
 {
 	if (nfs4_has_session(clp))
-		rpc_restart_call_prepare(task);
-	else
-		rpc_restart_call(task);
+		return rpc_restart_call_prepare(task);
+	return rpc_restart_call(task);
 }
diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c
index 81cf1425791..db8846a0e82 100644
--- a/fs/nfs/nfs2xdr.c
+++ b/fs/nfs/nfs2xdr.c
@@ -233,7 +233,7 @@ nfs_xdr_removeargs(struct rpc_rqst *req, __be32 *p, const struct nfs_removeargs
 static int
 nfs_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 	u32 offset = (u32)args->offset;
 	u32 count = args->count;
@@ -393,8 +393,7 @@ nfs_xdr_symlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs_symlinkargs *arg
 static int
 nfs_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs_readdirargs *args)
 {
-	struct rpc_task	*task = req->rq_task;
-	struct rpc_auth	*auth = task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 	u32 count = args->count;
 
@@ -575,7 +574,7 @@ nfs_xdr_diropres(struct rpc_rqst *req, __be32 *p, struct nfs_diropok *res)
 static int
 nfs_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs_readlinkargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 
 	p = xdr_encode_fhandle(p, args->fh);
diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c
index 75dcfc7da36..9769704f8ce 100644
--- a/fs/nfs/nfs3xdr.c
+++ b/fs/nfs/nfs3xdr.c
@@ -330,7 +330,7 @@ nfs3_xdr_accessargs(struct rpc_rqst *req, __be32 *p, struct nfs3_accessargs *arg
 static int
 nfs3_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 	u32 count = args->count;
 
@@ -471,7 +471,7 @@ nfs3_xdr_linkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_linkargs *args)
 static int
 nfs3_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 	u32 count = args->count;
 
@@ -675,7 +675,7 @@ static int
 nfs3_xdr_getaclargs(struct rpc_rqst *req, __be32 *p,
 		    struct nfs3_getaclargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 
 	p = xdr_encode_fhandle(p, args->fh);
@@ -802,7 +802,7 @@ nfs3_xdr_accessres(struct rpc_rqst *req, __be32 *p, struct nfs3_accessres *res)
 static int
 nfs3_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readlinkargs *args)
 {
-	struct rpc_auth	*auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth	*auth = req->rq_cred->cr_auth;
 	unsigned int replen;
 
 	p = xdr_encode_fhandle(p, args->fh);
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index c538c6106e1..311e15cc8af 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -45,10 +45,29 @@ enum nfs4_client_state {
 	NFS4CLNT_RECLAIM_NOGRACE,
 	NFS4CLNT_DELEGRETURN,
 	NFS4CLNT_SESSION_RESET,
-	NFS4CLNT_SESSION_DRAINING,
 	NFS4CLNT_RECALL_SLOT,
 };
 
+enum nfs4_session_state {
+	NFS4_SESSION_INITING,
+	NFS4_SESSION_DRAINING,
+};
+
+struct nfs4_minor_version_ops {
+	u32	minor_version;
+
+	int	(*call_sync)(struct nfs_server *server,
+			struct rpc_message *msg,
+			struct nfs4_sequence_args *args,
+			struct nfs4_sequence_res *res,
+			int cache_reply);
+	int	(*validate_stateid)(struct nfs_delegation *,
+			const nfs4_stateid *);
+	const struct nfs4_state_recovery_ops *reboot_recovery_ops;
+	const struct nfs4_state_recovery_ops *nograce_recovery_ops;
+	const struct nfs4_state_maintenance_ops *state_renewal_ops;
+};
+
 /*
  * struct rpc_sequence ensures that RPC calls are sent in the exact
  * order that they appear on the list.
@@ -89,7 +108,6 @@ struct nfs_unique_id {
  */
 struct nfs4_state_owner {
 	struct nfs_unique_id so_owner_id;
-	struct nfs_client    *so_client;
 	struct nfs_server    *so_server;
 	struct rb_node	     so_client_node;
 
@@ -99,7 +117,6 @@ struct nfs4_state_owner {
 	atomic_t	     so_count;
 	unsigned long	     so_flags;
 	struct list_head     so_states;
-	struct list_head     so_delegations;
 	struct nfs_seqid_counter so_seqid;
 	struct rpc_sequence  so_sequence;
 };
@@ -125,10 +142,20 @@ enum {
  * LOCK: one nfs4_state (LOCK) to hold the lock stateid nfs4_state(OPEN)
  */
 
+struct nfs4_lock_owner {
+	unsigned int lo_type;
+#define NFS4_ANY_LOCK_TYPE	(0U)
+#define NFS4_FLOCK_LOCK_TYPE	(1U << 0)
+#define NFS4_POSIX_LOCK_TYPE	(1U << 1)
+	union {
+		fl_owner_t posix_owner;
+		pid_t flock_owner;
+	} lo_u;
+};
+
 struct nfs4_lock_state {
 	struct list_head	ls_locks;	/* Other lock stateids */
 	struct nfs4_state *	ls_state;	/* Pointer to open state */
-	fl_owner_t		ls_owner;	/* POSIX lock owner */
 #define NFS_LOCK_INITIALIZED 1
 	int			ls_flags;
 	struct nfs_seqid_counter	ls_seqid;
@@ -136,6 +163,7 @@ struct nfs4_lock_state {
 	struct nfs_unique_id	ls_id;
 	nfs4_stateid		ls_stateid;
 	atomic_t		ls_count;
+	struct nfs4_lock_owner	ls_owner;
 };
 
 /* bits for nfs4_state->flags */
@@ -219,11 +247,15 @@ extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nam
 extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle);
 extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
 		struct nfs4_fs_locations *fs_locations, struct page *page);
+extern void nfs4_release_lockowner(const struct nfs4_lock_state *);
 
-extern struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[];
-extern struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[];
 #if defined(CONFIG_NFS_V4_1)
-extern int nfs4_setup_sequence(struct nfs_client *clp,
+static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server)
+{
+	return server->nfs_client->cl_session;
+}
+
+extern int nfs4_setup_sequence(const struct nfs_server *server,
 		struct nfs4_sequence_args *args, struct nfs4_sequence_res *res,
 		int cache_reply, struct rpc_task *task);
 extern void nfs4_destroy_session(struct nfs4_session *session);
@@ -234,7 +266,12 @@ extern int nfs4_init_session(struct nfs_server *server);
 extern int nfs4_proc_get_lease_time(struct nfs_client *clp,
 		struct nfs_fsinfo *fsinfo);
 #else /* CONFIG_NFS_v4_1 */
-static inline int nfs4_setup_sequence(struct nfs_client *clp,
+static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server)
+{
+	return NULL;
+}
+
+static inline int nfs4_setup_sequence(const struct nfs_server *server,
 		struct nfs4_sequence_args *args, struct nfs4_sequence_res *res,
 		int cache_reply, struct rpc_task *task)
 {
@@ -247,7 +284,7 @@ static inline int nfs4_init_session(struct nfs_server *server)
 }
 #endif /* CONFIG_NFS_V4_1 */
 
-extern struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[];
+extern const struct nfs4_minor_version_ops *nfs_v4_minor_ops[];
 
 extern const u32 nfs4_fattr_bitmap[2];
 extern const u32 nfs4_statfs_bitmap[2];
@@ -284,7 +321,7 @@ extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags)
 extern void nfs41_handle_recall_slot(struct nfs_client *clp);
 extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp);
 extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl);
-extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t);
+extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t, pid_t);
 
 extern struct nfs_seqid *nfs_alloc_seqid(struct nfs_seqid_counter *counter, gfp_t gfp_mask);
 extern int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 70015dd60a9..7ffbb98ddec 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -303,15 +303,19 @@ do_state_recovery:
 }
 
 
-static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
+static void do_renew_lease(struct nfs_client *clp, unsigned long timestamp)
 {
-	struct nfs_client *clp = server->nfs_client;
 	spin_lock(&clp->cl_lock);
 	if (time_before(clp->cl_last_renewal,timestamp))
 		clp->cl_last_renewal = timestamp;
 	spin_unlock(&clp->cl_lock);
 }
 
+static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
+{
+	do_renew_lease(server->nfs_client, timestamp);
+}
+
 #if defined(CONFIG_NFS_V4_1)
 
 /*
@@ -356,7 +360,7 @@ static void nfs41_check_drain_session_complete(struct nfs4_session *ses)
 {
 	struct rpc_task *task;
 
-	if (!test_bit(NFS4CLNT_SESSION_DRAINING, &ses->clp->cl_state)) {
+	if (!test_bit(NFS4_SESSION_DRAINING, &ses->session_state)) {
 		task = rpc_wake_up_next(&ses->fc_slot_table.slot_tbl_waitq);
 		if (task)
 			rpc_task_set_priority(task, RPC_PRIORITY_PRIVILEGED);
@@ -370,12 +374,11 @@ static void nfs41_check_drain_session_complete(struct nfs4_session *ses)
 	complete(&ses->complete);
 }
 
-static void nfs41_sequence_free_slot(const struct nfs_client *clp,
-			      struct nfs4_sequence_res *res)
+static void nfs41_sequence_free_slot(struct nfs4_sequence_res *res)
 {
 	struct nfs4_slot_table *tbl;
 
-	tbl = &clp->cl_session->fc_slot_table;
+	tbl = &res->sr_session->fc_slot_table;
 	if (res->sr_slotid == NFS4_MAX_SLOT_TABLE) {
 		/* just wake up the next guy waiting since
 		 * we may have not consumed a slot after all */
@@ -385,18 +388,17 @@ static void nfs41_sequence_free_slot(const struct nfs_client *clp,
 
 	spin_lock(&tbl->slot_tbl_lock);
 	nfs4_free_slot(tbl, res->sr_slotid);
-	nfs41_check_drain_session_complete(clp->cl_session);
+	nfs41_check_drain_session_complete(res->sr_session);
 	spin_unlock(&tbl->slot_tbl_lock);
 	res->sr_slotid = NFS4_MAX_SLOT_TABLE;
 }
 
-static void nfs41_sequence_done(struct nfs_client *clp,
-				struct nfs4_sequence_res *res,
-				int rpc_status)
+static int nfs41_sequence_done(struct rpc_task *task, struct nfs4_sequence_res *res)
 {
 	unsigned long timestamp;
 	struct nfs4_slot_table *tbl;
 	struct nfs4_slot *slot;
+	struct nfs_client *clp;
 
 	/*
 	 * sr_status remains 1 if an RPC level error occurred. The server
@@ -411,25 +413,51 @@ static void nfs41_sequence_done(struct nfs_client *clp,
 	if (res->sr_slotid == NFS4_MAX_SLOT_TABLE)
 		goto out;
 
+	tbl = &res->sr_session->fc_slot_table;
+	slot = tbl->slots + res->sr_slotid;
+
 	/* Check the SEQUENCE operation status */
-	if (res->sr_status == 0) {
-		tbl = &clp->cl_session->fc_slot_table;
-		slot = tbl->slots + res->sr_slotid;
+	switch (res->sr_status) {
+	case 0:
 		/* Update the slot's sequence and clientid lease timer */
 		++slot->seq_nr;
 		timestamp = res->sr_renewal_time;
-		spin_lock(&clp->cl_lock);
-		if (time_before(clp->cl_last_renewal, timestamp))
-			clp->cl_last_renewal = timestamp;
-		spin_unlock(&clp->cl_lock);
+		clp = res->sr_session->clp;
+		do_renew_lease(clp, timestamp);
 		/* Check sequence flags */
 		if (atomic_read(&clp->cl_count) > 1)
 			nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags);
+		break;
+	case -NFS4ERR_DELAY:
+		/* The server detected a resend of the RPC call and
+		 * returned NFS4ERR_DELAY as per Section 2.10.6.2
+		 * of RFC5661.
+		 */
+		dprintk("%s: slot=%d seq=%d: Operation in progress\n",
+				__func__, res->sr_slotid, slot->seq_nr);
+		goto out_retry;
+	default:
+		/* Just update the slot sequence no. */
+		++slot->seq_nr;
 	}
 out:
 	/* The session may be reset by one of the error handlers. */
 	dprintk("%s: Error %d free the slot \n", __func__, res->sr_status);
-	nfs41_sequence_free_slot(clp, res);
+	nfs41_sequence_free_slot(res);
+	return 1;
+out_retry:
+	if (!rpc_restart_call(task))
+		goto out;
+	rpc_delay(task, NFS4_POLL_RETRY_MAX);
+	return 0;
+}
+
+static int nfs4_sequence_done(struct rpc_task *task,
+			       struct nfs4_sequence_res *res)
+{
+	if (res->sr_session == NULL)
+		return 1;
+	return nfs41_sequence_done(task, res);
 }
 
 /*
@@ -480,12 +508,11 @@ static int nfs41_setup_sequence(struct nfs4_session *session,
 	if (res->sr_slotid != NFS4_MAX_SLOT_TABLE)
 		return 0;
 
-	memset(res, 0, sizeof(*res));
 	res->sr_slotid = NFS4_MAX_SLOT_TABLE;
 	tbl = &session->fc_slot_table;
 
 	spin_lock(&tbl->slot_tbl_lock);
-	if (test_bit(NFS4CLNT_SESSION_DRAINING, &session->clp->cl_state) &&
+	if (test_bit(NFS4_SESSION_DRAINING, &session->session_state) &&
 	    !rpc_task_has_priority(task, RPC_PRIORITY_PRIVILEGED)) {
 		/*
 		 * The state manager will wait until the slot table is empty.
@@ -525,6 +552,7 @@ static int nfs41_setup_sequence(struct nfs4_session *session,
 	res->sr_session = session;
 	res->sr_slotid = slotid;
 	res->sr_renewal_time = jiffies;
+	res->sr_status_flags = 0;
 	/*
 	 * sr_status is only set in decode_sequence, and so will remain
 	 * set to 1 if an rpc level failure occurs.
@@ -533,33 +561,33 @@ static int nfs41_setup_sequence(struct nfs4_session *session,
 	return 0;
 }
 
-int nfs4_setup_sequence(struct nfs_client *clp,
+int nfs4_setup_sequence(const struct nfs_server *server,
 			struct nfs4_sequence_args *args,
 			struct nfs4_sequence_res *res,
 			int cache_reply,
 			struct rpc_task *task)
 {
+	struct nfs4_session *session = nfs4_get_session(server);
 	int ret = 0;
 
+	if (session == NULL) {
+		args->sa_session = NULL;
+		res->sr_session = NULL;
+		goto out;
+	}
+
 	dprintk("--> %s clp %p session %p sr_slotid %d\n",
-		__func__, clp, clp->cl_session, res->sr_slotid);
+		__func__, session->clp, session, res->sr_slotid);
 
-	if (!nfs4_has_session(clp))
-		goto out;
-	ret = nfs41_setup_sequence(clp->cl_session, args, res, cache_reply,
+	ret = nfs41_setup_sequence(session, args, res, cache_reply,
 				   task);
-	if (ret && ret != -EAGAIN) {
-		/* terminate rpc task */
-		task->tk_status = ret;
-		task->tk_action = NULL;
-	}
 out:
 	dprintk("<-- %s status=%d\n", __func__, ret);
 	return ret;
 }
 
 struct nfs41_call_sync_data {
-	struct nfs_client *clp;
+	const struct nfs_server *seq_server;
 	struct nfs4_sequence_args *seq_args;
 	struct nfs4_sequence_res *seq_res;
 	int cache_reply;
@@ -569,9 +597,9 @@ static void nfs41_call_sync_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs41_call_sync_data *data = calldata;
 
-	dprintk("--> %s data->clp->cl_session %p\n", __func__,
-		data->clp->cl_session);
-	if (nfs4_setup_sequence(data->clp, data->seq_args,
+	dprintk("--> %s data->seq_server %p\n", __func__, data->seq_server);
+
+	if (nfs4_setup_sequence(data->seq_server, data->seq_args,
 				data->seq_res, data->cache_reply, task))
 		return;
 	rpc_call_start(task);
@@ -587,7 +615,7 @@ static void nfs41_call_sync_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs41_call_sync_data *data = calldata;
 
-	nfs41_sequence_done(data->clp, data->seq_res, task->tk_status);
+	nfs41_sequence_done(task, data->seq_res);
 }
 
 struct rpc_call_ops nfs41_call_sync_ops = {
@@ -600,8 +628,7 @@ struct rpc_call_ops nfs41_call_priv_sync_ops = {
 	.rpc_call_done = nfs41_call_sync_done,
 };
 
-static int nfs4_call_sync_sequence(struct nfs_client *clp,
-				   struct rpc_clnt *clnt,
+static int nfs4_call_sync_sequence(struct nfs_server *server,
 				   struct rpc_message *msg,
 				   struct nfs4_sequence_args *args,
 				   struct nfs4_sequence_res *res,
@@ -611,13 +638,13 @@ static int nfs4_call_sync_sequence(struct nfs_client *clp,
 	int ret;
 	struct rpc_task *task;
 	struct nfs41_call_sync_data data = {
-		.clp = clp,
+		.seq_server = server,
 		.seq_args = args,
 		.seq_res = res,
 		.cache_reply = cache_reply,
 	};
 	struct rpc_task_setup task_setup = {
-		.rpc_client = clnt,
+		.rpc_client = server->client,
 		.rpc_message = msg,
 		.callback_ops = &nfs41_call_sync_ops,
 		.callback_data = &data
@@ -642,10 +669,15 @@ int _nfs4_call_sync_session(struct nfs_server *server,
 			    struct nfs4_sequence_res *res,
 			    int cache_reply)
 {
-	return nfs4_call_sync_sequence(server->nfs_client, server->client,
-				       msg, args, res, cache_reply, 0);
+	return nfs4_call_sync_sequence(server, msg, args, res, cache_reply, 0);
 }
 
+#else
+static int nfs4_sequence_done(struct rpc_task *task,
+			       struct nfs4_sequence_res *res)
+{
+	return 1;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 int _nfs4_call_sync(struct nfs_server *server,
@@ -659,18 +691,9 @@ int _nfs4_call_sync(struct nfs_server *server,
 }
 
 #define nfs4_call_sync(server, msg, args, res, cache_reply) \
-	(server)->nfs_client->cl_call_sync((server), (msg), &(args)->seq_args, \
+	(server)->nfs_client->cl_mvops->call_sync((server), (msg), &(args)->seq_args, \
 			&(res)->seq_res, (cache_reply))
 
-static void nfs4_sequence_done(const struct nfs_server *server,
-			       struct nfs4_sequence_res *res, int rpc_status)
-{
-#ifdef CONFIG_NFS_V4_1
-	if (nfs4_has_session(server->nfs_client))
-		nfs41_sequence_done(server->nfs_client, res, rpc_status);
-#endif /* CONFIG_NFS_V4_1 */
-}
-
 static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo)
 {
 	struct nfs_inode *nfsi = NFS_I(dir);
@@ -745,19 +768,14 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.server = server;
 	p->o_arg.bitmask = server->attr_bitmask;
 	p->o_arg.claim = NFS4_OPEN_CLAIM_NULL;
-	if (flags & O_EXCL) {
-		if (nfs4_has_persistent_session(server->nfs_client)) {
-			/* GUARDED */
-			p->o_arg.u.attrs = &p->attrs;
-			memcpy(&p->attrs, attrs, sizeof(p->attrs));
-		} else { /* EXCLUSIVE4_1 */
-			u32 *s = (u32 *) p->o_arg.u.verifier.data;
-			s[0] = jiffies;
-			s[1] = current->pid;
-		}
-	} else if (flags & O_CREAT) {
+	if (flags & O_CREAT) {
+		u32 *s;
+
 		p->o_arg.u.attrs = &p->attrs;
 		memcpy(&p->attrs, attrs, sizeof(p->attrs));
+		s = (u32 *) p->o_arg.u.verifier.data;
+		s[0] = jiffies;
+		s[1] = current->pid;
 	}
 	p->c_arg.fh = &p->o_res.fh;
 	p->c_arg.stateid = &p->o_res.stateid;
@@ -1255,8 +1273,6 @@ static void nfs4_open_confirm_done(struct rpc_task *task, void *calldata)
 	struct nfs4_opendata *data = calldata;
 
 	data->rpc_status = task->tk_status;
-	if (RPC_ASSASSINATED(task))
-		return;
 	if (data->rpc_status == 0) {
 		memcpy(data->o_res.stateid.data, data->c_res.stateid.data,
 				sizeof(data->o_res.stateid.data));
@@ -1356,13 +1372,13 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
 	}
 	/* Update sequence id. */
 	data->o_arg.id = sp->so_owner_id.id;
-	data->o_arg.clientid = sp->so_client->cl_clientid;
+	data->o_arg.clientid = sp->so_server->nfs_client->cl_clientid;
 	if (data->o_arg.claim == NFS4_OPEN_CLAIM_PREVIOUS) {
 		task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_NOATTR];
 		nfs_copy_fh(&data->o_res.fh, data->o_arg.fh);
 	}
 	data->timestamp = jiffies;
-	if (nfs4_setup_sequence(data->o_arg.server->nfs_client,
+	if (nfs4_setup_sequence(data->o_arg.server,
 				&data->o_arg.seq_args,
 				&data->o_res.seq_res, 1, task))
 		return;
@@ -1385,11 +1401,9 @@ static void nfs4_open_done(struct rpc_task *task, void *calldata)
 
 	data->rpc_status = task->tk_status;
 
-	nfs4_sequence_done(data->o_arg.server, &data->o_res.seq_res,
-			task->tk_status);
-
-	if (RPC_ASSASSINATED(task))
+	if (!nfs4_sequence_done(task, &data->o_res.seq_res))
 		return;
+
 	if (task->tk_status == 0) {
 		switch (data->o_res.f_attr->mode & S_IFMT) {
 			case S_IFREG:
@@ -1773,7 +1787,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 	if (nfs4_copy_delegation_stateid(&arg.stateid, inode)) {
 		/* Use that stateid */
 	} else if (state != NULL) {
-		nfs4_copy_stateid(&arg.stateid, state, current->files);
+		nfs4_copy_stateid(&arg.stateid, state, current->files, current->tgid);
 	} else
 		memcpy(&arg.stateid, &zero_stateid, sizeof(arg.stateid));
 
@@ -1838,8 +1852,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data)
 	struct nfs4_state *state = calldata->state;
 	struct nfs_server *server = NFS_SERVER(calldata->inode);
 
-	nfs4_sequence_done(server, &calldata->res.seq_res, task->tk_status);
-	if (RPC_ASSASSINATED(task))
+	if (!nfs4_sequence_done(task, &calldata->res.seq_res))
 		return;
         /* hmm. we are done with the inode, and in the process of freeing
 	 * the state_owner. we keep this around to process errors
@@ -1903,7 +1916,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 
 	nfs_fattr_init(calldata->res.fattr);
 	calldata->timestamp = jiffies;
-	if (nfs4_setup_sequence((NFS_SERVER(calldata->inode))->nfs_client,
+	if (nfs4_setup_sequence(NFS_SERVER(calldata->inode),
 				&calldata->arg.seq_args, &calldata->res.seq_res,
 				1, task))
 		return;
@@ -2648,7 +2661,8 @@ static int nfs4_proc_unlink_done(struct rpc_task *task, struct inode *dir)
 {
 	struct nfs_removeres *res = task->tk_msg.rpc_resp;
 
-	nfs4_sequence_done(res->server, &res->seq_res, task->tk_status);
+	if (!nfs4_sequence_done(task, &res->seq_res))
+		return 0;
 	if (nfs4_async_handle_error(task, res->server, NULL) == -EAGAIN)
 		return 0;
 	update_changeattr(dir, &res->cinfo);
@@ -3093,7 +3107,8 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data)
 
 	dprintk("--> %s\n", __func__);
 
-	nfs4_sequence_done(server, &data->res.seq_res, task->tk_status);
+	if (!nfs4_sequence_done(task, &data->res.seq_res))
+		return -EAGAIN;
 
 	if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) {
 		nfs_restart_rpc(task, server->nfs_client);
@@ -3116,8 +3131,8 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data)
 {
 	struct inode *inode = data->inode;
 	
-	nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res,
-			   task->tk_status);
+	if (!nfs4_sequence_done(task, &data->res.seq_res))
+		return -EAGAIN;
 
 	if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) {
 		nfs_restart_rpc(task, NFS_SERVER(inode)->nfs_client);
@@ -3145,8 +3160,9 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data)
 {
 	struct inode *inode = data->inode;
 	
-	nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res,
-			   task->tk_status);
+	if (!nfs4_sequence_done(task, &data->res.seq_res))
+		return -EAGAIN;
+
 	if (nfs4_async_handle_error(task, NFS_SERVER(inode), NULL) == -EAGAIN) {
 		nfs_restart_rpc(task, NFS_SERVER(inode)->nfs_client);
 		return -EAGAIN;
@@ -3196,10 +3212,7 @@ static void nfs4_renew_done(struct rpc_task *task, void *calldata)
 			nfs4_schedule_state_recovery(clp);
 		return;
 	}
-	spin_lock(&clp->cl_lock);
-	if (time_before(clp->cl_last_renewal,timestamp))
-		clp->cl_last_renewal = timestamp;
-	spin_unlock(&clp->cl_lock);
+	do_renew_lease(clp, timestamp);
 }
 
 static const struct rpc_call_ops nfs4_renew_ops = {
@@ -3240,10 +3253,7 @@ int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
 	status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
 	if (status < 0)
 		return status;
-	spin_lock(&clp->cl_lock);
-	if (time_before(clp->cl_last_renewal,now))
-		clp->cl_last_renewal = now;
-	spin_unlock(&clp->cl_lock);
+	do_renew_lease(clp, now);
 	return 0;
 }
 
@@ -3464,9 +3474,11 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen
 }
 
 static int
-_nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs_client *clp, struct nfs4_state *state)
+nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state)
 {
-	if (!clp || task->tk_status >= 0)
+	struct nfs_client *clp = server->nfs_client;
+
+	if (task->tk_status >= 0)
 		return 0;
 	switch(task->tk_status) {
 		case -NFS4ERR_ADMIN_REVOKED:
@@ -3498,8 +3510,7 @@ _nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server,
 			return -EAGAIN;
 #endif /* CONFIG_NFS_V4_1 */
 		case -NFS4ERR_DELAY:
-			if (server)
-				nfs_inc_server_stats(server, NFSIOS_DELAY);
+			nfs_inc_server_stats(server, NFSIOS_DELAY);
 		case -NFS4ERR_GRACE:
 		case -EKEYEXPIRED:
 			rpc_delay(task, NFS4_POLL_RETRY_MAX);
@@ -3520,12 +3531,6 @@ do_state_recovery:
 	return -EAGAIN;
 }
 
-static int
-nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state)
-{
-	return _nfs4_async_handle_error(task, server, server->nfs_client, state);
-}
-
 int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 		unsigned short port, struct rpc_cred *cred,
 		struct nfs4_setclientid_res *res)
@@ -3641,8 +3646,8 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs4_delegreturndata *data = calldata;
 
-	nfs4_sequence_done(data->res.server, &data->res.seq_res,
-			task->tk_status);
+	if (!nfs4_sequence_done(task, &data->res.seq_res))
+		return;
 
 	switch (task->tk_status) {
 	case -NFS4ERR_STALE_STATEID:
@@ -3672,7 +3677,7 @@ static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data)
 
 	d_data = (struct nfs4_delegreturndata *)data;
 
-	if (nfs4_setup_sequence(d_data->res.server->nfs_client,
+	if (nfs4_setup_sequence(d_data->res.server,
 				&d_data->args.seq_args,
 				&d_data->res.seq_res, 1, task))
 		return;
@@ -3892,9 +3897,7 @@ static void nfs4_locku_done(struct rpc_task *task, void *data)
 {
 	struct nfs4_unlockdata *calldata = data;
 
-	nfs4_sequence_done(calldata->server, &calldata->res.seq_res,
-			   task->tk_status);
-	if (RPC_ASSASSINATED(task))
+	if (!nfs4_sequence_done(task, &calldata->res.seq_res))
 		return;
 	switch (task->tk_status) {
 		case 0:
@@ -3927,7 +3930,7 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data)
 		return;
 	}
 	calldata->timestamp = jiffies;
-	if (nfs4_setup_sequence(calldata->server->nfs_client,
+	if (nfs4_setup_sequence(calldata->server,
 				&calldata->arg.seq_args,
 				&calldata->res.seq_res, 1, task))
 		return;
@@ -4082,7 +4085,8 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata)
 	} else
 		data->arg.new_lock_owner = 0;
 	data->timestamp = jiffies;
-	if (nfs4_setup_sequence(data->server->nfs_client, &data->arg.seq_args,
+	if (nfs4_setup_sequence(data->server,
+				&data->arg.seq_args,
 				&data->res.seq_res, 1, task))
 		return;
 	rpc_call_start(task);
@@ -4101,12 +4105,10 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata)
 
 	dprintk("%s: begin!\n", __func__);
 
-	nfs4_sequence_done(data->server, &data->res.seq_res,
-			task->tk_status);
+	if (!nfs4_sequence_done(task, &data->res.seq_res))
+		return;
 
 	data->rpc_status = task->tk_status;
-	if (RPC_ASSASSINATED(task))
-		goto out;
 	if (data->arg.new_lock_owner != 0) {
 		if (data->rpc_status == 0)
 			nfs_confirm_seqid(&data->lsp->ls_seqid, 0);
@@ -4424,6 +4426,34 @@ out:
 	return err;
 }
 
+static void nfs4_release_lockowner_release(void *calldata)
+{
+	kfree(calldata);
+}
+
+const struct rpc_call_ops nfs4_release_lockowner_ops = {
+	.rpc_release = nfs4_release_lockowner_release,
+};
+
+void nfs4_release_lockowner(const struct nfs4_lock_state *lsp)
+{
+	struct nfs_server *server = lsp->ls_state->owner->so_server;
+	struct nfs_release_lockowner_args *args;
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RELEASE_LOCKOWNER],
+	};
+
+	if (server->nfs_client->cl_mvops->minor_version != 0)
+		return;
+	args = kmalloc(sizeof(*args), GFP_NOFS);
+	if (!args)
+		return;
+	args->lock_owner.clientid = server->nfs_client->cl_clientid;
+	args->lock_owner.id = lsp->ls_id.id;
+	msg.rpc_argp = args;
+	rpc_call_async(server->client, &msg, 0, &nfs4_release_lockowner_ops, args);
+}
+
 #define XATTR_NAME_NFSV4_ACL "system.nfs4_acl"
 
 int nfs4_setxattr(struct dentry *dentry, const char *key, const void *buf,
@@ -4611,7 +4641,8 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata)
 			(struct nfs4_get_lease_time_data *)calldata;
 
 	dprintk("--> %s\n", __func__);
-	nfs41_sequence_done(data->clp, &data->res->lr_seq_res, task->tk_status);
+	if (!nfs41_sequence_done(task, &data->res->lr_seq_res))
+		return;
 	switch (task->tk_status) {
 	case -NFS4ERR_DELAY:
 	case -NFS4ERR_GRACE:
@@ -4805,13 +4836,6 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp)
 	if (!session)
 		return NULL;
 
-	/*
-	 * The create session reply races with the server back
-	 * channel probe. Mark the client NFS_CS_SESSION_INITING
-	 * so that the client back channel can find the
-	 * nfs_client struct
-	 */
-	clp->cl_cons_state = NFS_CS_SESSION_INITING;
 	init_completion(&session->complete);
 
 	tbl = &session->fc_slot_table;
@@ -4824,6 +4848,8 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp)
 	spin_lock_init(&tbl->slot_tbl_lock);
 	rpc_init_wait_queue(&tbl->slot_tbl_waitq, "BackChannel Slot table");
 
+	session->session_state = 1<<NFS4_SESSION_INITING;
+
 	session->clp = clp;
 	return session;
 }
@@ -5040,6 +5066,10 @@ int nfs4_init_session(struct nfs_server *server)
 	if (!nfs4_has_session(clp))
 		return 0;
 
+	session = clp->cl_session;
+	if (!test_and_clear_bit(NFS4_SESSION_INITING, &session->session_state))
+		return 0;
+
 	rsize = server->rsize;
 	if (rsize == 0)
 		rsize = NFS_MAX_FILE_IO_SIZE;
@@ -5047,7 +5077,6 @@ int nfs4_init_session(struct nfs_server *server)
 	if (wsize == 0)
 		wsize = NFS_MAX_FILE_IO_SIZE;
 
-	session = clp->cl_session;
 	session->fc_attrs.max_rqst_sz = wsize + nfs41_maxwrite_overhead;
 	session->fc_attrs.max_resp_sz = rsize + nfs41_maxread_overhead;
 
@@ -5060,69 +5089,70 @@ int nfs4_init_session(struct nfs_server *server)
 /*
  * Renew the cl_session lease.
  */
-static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred)
-{
+struct nfs4_sequence_data {
+	struct nfs_client *clp;
 	struct nfs4_sequence_args args;
 	struct nfs4_sequence_res res;
-
-	struct rpc_message msg = {
-		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE],
-		.rpc_argp = &args,
-		.rpc_resp = &res,
-		.rpc_cred = cred,
-	};
-
-	args.sa_cache_this = 0;
-
-	return nfs4_call_sync_sequence(clp, clp->cl_rpcclient, &msg, &args,
-				       &res, args.sa_cache_this, 1);
-}
+};
 
 static void nfs41_sequence_release(void *data)
 {
-	struct nfs_client *clp = (struct nfs_client *)data;
+	struct nfs4_sequence_data *calldata = data;
+	struct nfs_client *clp = calldata->clp;
 
 	if (atomic_read(&clp->cl_count) > 1)
 		nfs4_schedule_state_renewal(clp);
 	nfs_put_client(clp);
+	kfree(calldata);
+}
+
+static int nfs41_sequence_handle_errors(struct rpc_task *task, struct nfs_client *clp)
+{
+	switch(task->tk_status) {
+	case -NFS4ERR_DELAY:
+	case -EKEYEXPIRED:
+		rpc_delay(task, NFS4_POLL_RETRY_MAX);
+		return -EAGAIN;
+	default:
+		nfs4_schedule_state_recovery(clp);
+	}
+	return 0;
 }
 
 static void nfs41_sequence_call_done(struct rpc_task *task, void *data)
 {
-	struct nfs_client *clp = (struct nfs_client *)data;
+	struct nfs4_sequence_data *calldata = data;
+	struct nfs_client *clp = calldata->clp;
 
-	nfs41_sequence_done(clp, task->tk_msg.rpc_resp, task->tk_status);
+	if (!nfs41_sequence_done(task, task->tk_msg.rpc_resp))
+		return;
 
 	if (task->tk_status < 0) {
 		dprintk("%s ERROR %d\n", __func__, task->tk_status);
 		if (atomic_read(&clp->cl_count) == 1)
 			goto out;
 
-		if (_nfs4_async_handle_error(task, NULL, clp, NULL)
-								== -EAGAIN) {
-			nfs_restart_rpc(task, clp);
+		if (nfs41_sequence_handle_errors(task, clp) == -EAGAIN) {
+			rpc_restart_call_prepare(task);
 			return;
 		}
 	}
 	dprintk("%s rpc_cred %p\n", __func__, task->tk_msg.rpc_cred);
 out:
-	kfree(task->tk_msg.rpc_argp);
-	kfree(task->tk_msg.rpc_resp);
-
 	dprintk("<-- %s\n", __func__);
 }
 
 static void nfs41_sequence_prepare(struct rpc_task *task, void *data)
 {
-	struct nfs_client *clp;
+	struct nfs4_sequence_data *calldata = data;
+	struct nfs_client *clp = calldata->clp;
 	struct nfs4_sequence_args *args;
 	struct nfs4_sequence_res *res;
 
-	clp = (struct nfs_client *)data;
 	args = task->tk_msg.rpc_argp;
 	res = task->tk_msg.rpc_resp;
 
-	if (nfs4_setup_sequence(clp, args, res, 0, task))
+	if (nfs41_setup_sequence(clp->cl_session, args, res, 0, task))
 		return;
 	rpc_call_start(task);
 }
@@ -5133,32 +5163,67 @@ static const struct rpc_call_ops nfs41_sequence_ops = {
 	.rpc_release = nfs41_sequence_release,
 };
 
-static int nfs41_proc_async_sequence(struct nfs_client *clp,
-				     struct rpc_cred *cred)
+static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred)
 {
-	struct nfs4_sequence_args *args;
-	struct nfs4_sequence_res *res;
+	struct nfs4_sequence_data *calldata;
 	struct rpc_message msg = {
 		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE],
 		.rpc_cred = cred,
 	};
+	struct rpc_task_setup task_setup_data = {
+		.rpc_client = clp->cl_rpcclient,
+		.rpc_message = &msg,
+		.callback_ops = &nfs41_sequence_ops,
+		.flags = RPC_TASK_ASYNC | RPC_TASK_SOFT,
+	};
 
 	if (!atomic_inc_not_zero(&clp->cl_count))
-		return -EIO;
-	args = kzalloc(sizeof(*args), GFP_NOFS);
-	res = kzalloc(sizeof(*res), GFP_NOFS);
-	if (!args || !res) {
-		kfree(args);
-		kfree(res);
+		return ERR_PTR(-EIO);
+	calldata = kmalloc(sizeof(*calldata), GFP_NOFS);
+	if (calldata == NULL) {
 		nfs_put_client(clp);
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 	}
-	res->sr_slotid = NFS4_MAX_SLOT_TABLE;
-	msg.rpc_argp = args;
-	msg.rpc_resp = res;
+	calldata->res.sr_slotid = NFS4_MAX_SLOT_TABLE;
+	msg.rpc_argp = &calldata->args;
+	msg.rpc_resp = &calldata->res;
+	calldata->clp = clp;
+	task_setup_data.callback_data = calldata;
 
-	return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_SOFT,
-			      &nfs41_sequence_ops, (void *)clp);
+	return rpc_run_task(&task_setup_data);
+}
+
+static int nfs41_proc_async_sequence(struct nfs_client *clp, struct rpc_cred *cred)
+{
+	struct rpc_task *task;
+	int ret = 0;
+
+	task = _nfs41_proc_sequence(clp, cred);
+	if (IS_ERR(task))
+		ret = PTR_ERR(task);
+	else
+		rpc_put_task(task);
+	dprintk("<-- %s status=%d\n", __func__, ret);
+	return ret;
+}
+
+static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred)
+{
+	struct rpc_task *task;
+	int ret;
+
+	task = _nfs41_proc_sequence(clp, cred);
+	if (IS_ERR(task)) {
+		ret = PTR_ERR(task);
+		goto out;
+	}
+	ret = rpc_wait_for_completion_task(task);
+	if (!ret)
+		ret = task->tk_status;
+	rpc_put_task(task);
+out:
+	dprintk("<-- %s status=%d\n", __func__, ret);
+	return ret;
 }
 
 struct nfs4_reclaim_complete_data {
@@ -5172,13 +5237,31 @@ static void nfs4_reclaim_complete_prepare(struct rpc_task *task, void *data)
 	struct nfs4_reclaim_complete_data *calldata = data;
 
 	rpc_task_set_priority(task, RPC_PRIORITY_PRIVILEGED);
-	if (nfs4_setup_sequence(calldata->clp, &calldata->arg.seq_args,
+	if (nfs41_setup_sequence(calldata->clp->cl_session,
+				&calldata->arg.seq_args,
 				&calldata->res.seq_res, 0, task))
 		return;
 
 	rpc_call_start(task);
 }
 
+static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nfs_client *clp)
+{
+	switch(task->tk_status) {
+	case 0:
+	case -NFS4ERR_COMPLETE_ALREADY:
+	case -NFS4ERR_WRONG_CRED: /* What to do here? */
+		break;
+	case -NFS4ERR_DELAY:
+	case -EKEYEXPIRED:
+		rpc_delay(task, NFS4_POLL_RETRY_MAX);
+		return -EAGAIN;
+	default:
+		nfs4_schedule_state_recovery(clp);
+	}
+	return 0;
+}
+
 static void nfs4_reclaim_complete_done(struct rpc_task *task, void *data)
 {
 	struct nfs4_reclaim_complete_data *calldata = data;
@@ -5186,32 +5269,13 @@ static void nfs4_reclaim_complete_done(struct rpc_task *task, void *data)
 	struct nfs4_sequence_res *res = &calldata->res.seq_res;
 
 	dprintk("--> %s\n", __func__);
-	nfs41_sequence_done(clp, res, task->tk_status);
-	switch (task->tk_status) {
-	case 0:
-	case -NFS4ERR_COMPLETE_ALREADY:
-		break;
-	case -NFS4ERR_BADSESSION:
-	case -NFS4ERR_DEADSESSION:
-		/*
-		 * Handle the session error, but do not retry the operation, as
-		 * we have no way of telling whether the clientid had to be
-		 * reset before we got our reply.  If reset, a new wave of
-		 * reclaim operations will follow, containing their own reclaim
-		 * complete.  We don't want our retry to get on the way of
-		 * recovery by incorrectly indicating to the server that we're
-		 * done reclaiming state since the process had to be restarted.
-		 */
-		_nfs4_async_handle_error(task, NULL, clp, NULL);
-		break;
-	default:
-		if (_nfs4_async_handle_error(
-				task, NULL, clp, NULL) == -EAGAIN) {
-			rpc_restart_call_prepare(task);
-			return;
-		}
-	}
+	if (!nfs41_sequence_done(task, res))
+		return;
 
+	if (nfs41_reclaim_complete_handle_errors(task, clp) == -EAGAIN) {
+		rpc_restart_call_prepare(task);
+		return;
+	}
 	dprintk("<-- %s\n", __func__);
 }
 
@@ -5325,28 +5389,30 @@ struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = {
 };
 #endif
 
-/*
- * Per minor version reboot and network partition recovery ops
- */
-
-struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[] = {
-	&nfs40_reboot_recovery_ops,
-#if defined(CONFIG_NFS_V4_1)
-	&nfs41_reboot_recovery_ops,
-#endif
+static const struct nfs4_minor_version_ops nfs_v4_0_minor_ops = {
+	.minor_version = 0,
+	.call_sync = _nfs4_call_sync,
+	.validate_stateid = nfs4_validate_delegation_stateid,
+	.reboot_recovery_ops = &nfs40_reboot_recovery_ops,
+	.nograce_recovery_ops = &nfs40_nograce_recovery_ops,
+	.state_renewal_ops = &nfs40_state_renewal_ops,
 };
 
-struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[] = {
-	&nfs40_nograce_recovery_ops,
 #if defined(CONFIG_NFS_V4_1)
-	&nfs41_nograce_recovery_ops,
-#endif
+static const struct nfs4_minor_version_ops nfs_v4_1_minor_ops = {
+	.minor_version = 1,
+	.call_sync = _nfs4_call_sync_session,
+	.validate_stateid = nfs41_validate_delegation_stateid,
+	.reboot_recovery_ops = &nfs41_reboot_recovery_ops,
+	.nograce_recovery_ops = &nfs41_nograce_recovery_ops,
+	.state_renewal_ops = &nfs41_state_renewal_ops,
 };
+#endif
 
-struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[] = {
-	&nfs40_state_renewal_ops,
+const struct nfs4_minor_version_ops *nfs_v4_minor_ops[] = {
+	[0] = &nfs_v4_0_minor_ops,
 #if defined(CONFIG_NFS_V4_1)
-	&nfs41_state_renewal_ops,
+	[1] = &nfs_v4_1_minor_ops,
 #endif
 };
 
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index d87f10327b7..72b6c580af1 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -54,14 +54,14 @@
 void
 nfs4_renew_state(struct work_struct *work)
 {
-	struct nfs4_state_maintenance_ops *ops;
+	const struct nfs4_state_maintenance_ops *ops;
 	struct nfs_client *clp =
 		container_of(work, struct nfs_client, cl_renewd.work);
 	struct rpc_cred *cred;
 	long lease;
 	unsigned long last, now;
 
-	ops = nfs4_state_renewal_ops[clp->cl_minorversion];
+	ops = clp->cl_mvops->state_renewal_ops;
 	dprintk("%s: start\n", __func__);
 	/* Are there any active superblocks? */
 	if (list_empty(&clp->cl_superblocks))
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 34acf5926fd..3e2f19b04c0 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -145,7 +145,9 @@ static void nfs4_end_drain_session(struct nfs_client *clp)
 	struct nfs4_session *ses = clp->cl_session;
 	int max_slots;
 
-	if (test_and_clear_bit(NFS4CLNT_SESSION_DRAINING, &clp->cl_state)) {
+	if (ses == NULL)
+		return;
+	if (test_and_clear_bit(NFS4_SESSION_DRAINING, &ses->session_state)) {
 		spin_lock(&ses->fc_slot_table.slot_tbl_lock);
 		max_slots = ses->fc_slot_table.max_slots;
 		while (max_slots--) {
@@ -167,7 +169,7 @@ static int nfs4_begin_drain_session(struct nfs_client *clp)
 	struct nfs4_slot_table *tbl = &ses->fc_slot_table;
 
 	spin_lock(&tbl->slot_tbl_lock);
-	set_bit(NFS4CLNT_SESSION_DRAINING, &clp->cl_state);
+	set_bit(NFS4_SESSION_DRAINING, &ses->session_state);
 	if (tbl->highest_used_slotid != -1) {
 		INIT_COMPLETION(ses->complete);
 		spin_unlock(&tbl->slot_tbl_lock);
@@ -371,7 +373,6 @@ nfs4_alloc_state_owner(void)
 		return NULL;
 	spin_lock_init(&sp->so_lock);
 	INIT_LIST_HEAD(&sp->so_states);
-	INIT_LIST_HEAD(&sp->so_delegations);
 	rpc_init_wait_queue(&sp->so_sequence.wait, "Seqid_waitqueue");
 	sp->so_seqid.sequence = &sp->so_sequence;
 	spin_lock_init(&sp->so_sequence.lock);
@@ -384,7 +385,7 @@ static void
 nfs4_drop_state_owner(struct nfs4_state_owner *sp)
 {
 	if (!RB_EMPTY_NODE(&sp->so_client_node)) {
-		struct nfs_client *clp = sp->so_client;
+		struct nfs_client *clp = sp->so_server->nfs_client;
 
 		spin_lock(&clp->cl_lock);
 		rb_erase(&sp->so_client_node, &clp->cl_state_owners);
@@ -406,7 +407,6 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct
 	new = nfs4_alloc_state_owner();
 	if (new == NULL)
 		return NULL;
-	new->so_client = clp;
 	new->so_server = server;
 	new->so_cred = cred;
 	spin_lock(&clp->cl_lock);
@@ -423,7 +423,7 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct
 
 void nfs4_put_state_owner(struct nfs4_state_owner *sp)
 {
-	struct nfs_client *clp = sp->so_client;
+	struct nfs_client *clp = sp->so_server->nfs_client;
 	struct rpc_cred *cred = sp->so_cred;
 
 	if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock))
@@ -602,12 +602,21 @@ void nfs4_close_sync(struct path *path, struct nfs4_state *state, fmode_t fmode)
  * that is compatible with current->files
  */
 static struct nfs4_lock_state *
-__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
+__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type)
 {
 	struct nfs4_lock_state *pos;
 	list_for_each_entry(pos, &state->lock_states, ls_locks) {
-		if (pos->ls_owner != fl_owner)
+		if (type != NFS4_ANY_LOCK_TYPE && pos->ls_owner.lo_type != type)
 			continue;
+		switch (pos->ls_owner.lo_type) {
+		case NFS4_POSIX_LOCK_TYPE:
+			if (pos->ls_owner.lo_u.posix_owner != fl_owner)
+				continue;
+			break;
+		case NFS4_FLOCK_LOCK_TYPE:
+			if (pos->ls_owner.lo_u.flock_owner != fl_pid)
+				continue;
+		}
 		atomic_inc(&pos->ls_count);
 		return pos;
 	}
@@ -619,10 +628,10 @@ __nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
  * exists, return an uninitialized one.
  *
  */
-static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
+static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type)
 {
 	struct nfs4_lock_state *lsp;
-	struct nfs_client *clp = state->owner->so_client;
+	struct nfs_client *clp = state->owner->so_server->nfs_client;
 
 	lsp = kzalloc(sizeof(*lsp), GFP_NOFS);
 	if (lsp == NULL)
@@ -633,7 +642,18 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f
 	lsp->ls_seqid.sequence = &lsp->ls_sequence;
 	atomic_set(&lsp->ls_count, 1);
 	lsp->ls_state = state;
-	lsp->ls_owner = fl_owner;
+	lsp->ls_owner.lo_type = type;
+	switch (lsp->ls_owner.lo_type) {
+	case NFS4_FLOCK_LOCK_TYPE:
+		lsp->ls_owner.lo_u.flock_owner = fl_pid;
+		break;
+	case NFS4_POSIX_LOCK_TYPE:
+		lsp->ls_owner.lo_u.posix_owner = fl_owner;
+		break;
+	default:
+		kfree(lsp);
+		return NULL;
+	}
 	spin_lock(&clp->cl_lock);
 	nfs_alloc_unique_id(&clp->cl_lockowner_id, &lsp->ls_id, 1, 64);
 	spin_unlock(&clp->cl_lock);
@@ -643,7 +663,7 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f
 
 static void nfs4_free_lock_state(struct nfs4_lock_state *lsp)
 {
-	struct nfs_client *clp = lsp->ls_state->owner->so_client;
+	struct nfs_client *clp = lsp->ls_state->owner->so_server->nfs_client;
 
 	spin_lock(&clp->cl_lock);
 	nfs_free_unique_id(&clp->cl_lockowner_id, &lsp->ls_id);
@@ -657,13 +677,13 @@ static void nfs4_free_lock_state(struct nfs4_lock_state *lsp)
  * exists, return an uninitialized one.
  *
  */
-static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner)
+static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner, pid_t pid, unsigned int type)
 {
 	struct nfs4_lock_state *lsp, *new = NULL;
 	
 	for(;;) {
 		spin_lock(&state->state_lock);
-		lsp = __nfs4_find_lock_state(state, owner);
+		lsp = __nfs4_find_lock_state(state, owner, pid, type);
 		if (lsp != NULL)
 			break;
 		if (new != NULL) {
@@ -674,7 +694,7 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_
 			break;
 		}
 		spin_unlock(&state->state_lock);
-		new = nfs4_alloc_lock_state(state, owner);
+		new = nfs4_alloc_lock_state(state, owner, pid, type);
 		if (new == NULL)
 			return NULL;
 	}
@@ -701,6 +721,8 @@ void nfs4_put_lock_state(struct nfs4_lock_state *lsp)
 	if (list_empty(&state->lock_states))
 		clear_bit(LK_STATE_IN_USE, &state->flags);
 	spin_unlock(&state->state_lock);
+	if (lsp->ls_flags & NFS_LOCK_INITIALIZED)
+		nfs4_release_lockowner(lsp);
 	nfs4_free_lock_state(lsp);
 }
 
@@ -728,7 +750,12 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl)
 
 	if (fl->fl_ops != NULL)
 		return 0;
-	lsp = nfs4_get_lock_state(state, fl->fl_owner);
+	if (fl->fl_flags & FL_POSIX)
+		lsp = nfs4_get_lock_state(state, fl->fl_owner, 0, NFS4_POSIX_LOCK_TYPE);
+	else if (fl->fl_flags & FL_FLOCK)
+		lsp = nfs4_get_lock_state(state, 0, fl->fl_pid, NFS4_FLOCK_LOCK_TYPE);
+	else
+		return -EINVAL;
 	if (lsp == NULL)
 		return -ENOMEM;
 	fl->fl_u.nfs4_fl.owner = lsp;
@@ -740,7 +767,7 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl)
  * Byte-range lock aware utility to initialize the stateid of read/write
  * requests.
  */
-void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t fl_owner)
+void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid)
 {
 	struct nfs4_lock_state *lsp;
 	int seq;
@@ -753,7 +780,7 @@ void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t f
 		return;
 
 	spin_lock(&state->state_lock);
-	lsp = __nfs4_find_lock_state(state, fl_owner);
+	lsp = __nfs4_find_lock_state(state, fl_owner, fl_pid, NFS4_ANY_LOCK_TYPE);
 	if (lsp != NULL && (lsp->ls_flags & NFS_LOCK_INITIALIZED) != 0)
 		memcpy(dst, &lsp->ls_stateid, sizeof(*dst));
 	spin_unlock(&state->state_lock);
@@ -1041,11 +1068,11 @@ restart:
 			case -NFS4ERR_BAD_STATEID:
 			case -NFS4ERR_RECLAIM_BAD:
 			case -NFS4ERR_RECLAIM_CONFLICT:
-				nfs4_state_mark_reclaim_nograce(sp->so_client, state);
+				nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state);
 				break;
 			case -NFS4ERR_EXPIRED:
 			case -NFS4ERR_NO_GRACE:
-				nfs4_state_mark_reclaim_nograce(sp->so_client, state);
+				nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state);
 			case -NFS4ERR_STALE_CLIENTID:
 			case -NFS4ERR_BADSESSION:
 			case -NFS4ERR_BADSLOT:
@@ -1120,8 +1147,7 @@ static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp)
 	if (!test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state))
 		return;
 
-	nfs4_reclaim_complete(clp,
-		nfs4_reboot_recovery_ops[clp->cl_minorversion]);
+	nfs4_reclaim_complete(clp, clp->cl_mvops->reboot_recovery_ops);
 
 	for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) {
 		sp = rb_entry(pos, struct nfs4_state_owner, so_client_node);
@@ -1211,8 +1237,8 @@ restart:
 static int nfs4_check_lease(struct nfs_client *clp)
 {
 	struct rpc_cred *cred;
-	struct nfs4_state_maintenance_ops *ops =
-		nfs4_state_renewal_ops[clp->cl_minorversion];
+	const struct nfs4_state_maintenance_ops *ops =
+		clp->cl_mvops->state_renewal_ops;
 	int status = -NFS4ERR_EXPIRED;
 
 	/* Is the client already known to have an expired lease? */
@@ -1235,8 +1261,8 @@ out:
 static int nfs4_reclaim_lease(struct nfs_client *clp)
 {
 	struct rpc_cred *cred;
-	struct nfs4_state_recovery_ops *ops =
-		nfs4_reboot_recovery_ops[clp->cl_minorversion];
+	const struct nfs4_state_recovery_ops *ops =
+		clp->cl_mvops->reboot_recovery_ops;
 	int status = -ENOENT;
 
 	cred = ops->get_clid_cred(clp);
@@ -1444,7 +1470,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
 		/* First recover reboot state... */
 		if (test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) {
 			status = nfs4_do_reclaim(clp,
-				nfs4_reboot_recovery_ops[clp->cl_minorversion]);
+				clp->cl_mvops->reboot_recovery_ops);
 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) ||
 			    test_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state))
 				continue;
@@ -1458,7 +1484,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
 		/* Now recover expired state... */
 		if (test_and_clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state)) {
 			status = nfs4_do_reclaim(clp,
-				nfs4_nograce_recovery_ops[clp->cl_minorversion]);
+				clp->cl_mvops->nograce_recovery_ops);
 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) ||
 			    test_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state) ||
 			    test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state))
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 65c8dae4b26..08ef9129113 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -202,14 +202,17 @@ static int nfs4_stat_to_errno(int);
 #define encode_link_maxsz	(op_encode_hdr_maxsz + \
 				nfs4_name_maxsz)
 #define decode_link_maxsz	(op_decode_hdr_maxsz + decode_change_info_maxsz)
+#define encode_lockowner_maxsz	(7)
 #define encode_lock_maxsz	(op_encode_hdr_maxsz + \
 				 7 + \
-				 1 + encode_stateid_maxsz + 8)
+				 1 + encode_stateid_maxsz + 1 + \
+				 encode_lockowner_maxsz)
 #define decode_lock_denied_maxsz \
 				(8 + decode_lockowner_maxsz)
 #define decode_lock_maxsz	(op_decode_hdr_maxsz + \
 				 decode_lock_denied_maxsz)
-#define encode_lockt_maxsz	(op_encode_hdr_maxsz + 12)
+#define encode_lockt_maxsz	(op_encode_hdr_maxsz + 5 + \
+				encode_lockowner_maxsz)
 #define decode_lockt_maxsz	(op_decode_hdr_maxsz + \
 				 decode_lock_denied_maxsz)
 #define encode_locku_maxsz	(op_encode_hdr_maxsz + 3 + \
@@ -217,6 +220,11 @@ static int nfs4_stat_to_errno(int);
 				 4)
 #define decode_locku_maxsz	(op_decode_hdr_maxsz + \
 				 decode_stateid_maxsz)
+#define encode_release_lockowner_maxsz \
+				(op_encode_hdr_maxsz + \
+				 encode_lockowner_maxsz)
+#define decode_release_lockowner_maxsz \
+				(op_decode_hdr_maxsz)
 #define encode_access_maxsz	(op_encode_hdr_maxsz + 1)
 #define decode_access_maxsz	(op_decode_hdr_maxsz + 2)
 #define encode_symlink_maxsz	(op_encode_hdr_maxsz + \
@@ -471,6 +479,12 @@ static int nfs4_stat_to_errno(int);
 				decode_sequence_maxsz + \
 				decode_putfh_maxsz + \
 				decode_locku_maxsz)
+#define NFS4_enc_release_lockowner_sz \
+				(compound_encode_hdr_maxsz + \
+				 encode_lockowner_maxsz)
+#define NFS4_dec_release_lockowner_sz \
+				(compound_decode_hdr_maxsz + \
+				 decode_lockowner_maxsz)
 #define NFS4_enc_access_sz	(compound_encode_hdr_maxsz + \
 				encode_sequence_maxsz + \
 				encode_putfh_maxsz + \
@@ -744,7 +758,7 @@ static void encode_compound_hdr(struct xdr_stream *xdr,
 				struct compound_hdr *hdr)
 {
 	__be32 *p;
-	struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth;
+	struct rpc_auth *auth = req->rq_cred->cr_auth;
 
 	/* initialize running count of expected bytes in reply.
 	 * NOTE: the replied tag SHOULD be the same is the one sent,
@@ -1042,6 +1056,17 @@ static inline uint64_t nfs4_lock_length(struct file_lock *fl)
 	return fl->fl_end - fl->fl_start + 1;
 }
 
+static void encode_lockowner(struct xdr_stream *xdr, const struct nfs_lowner *lowner)
+{
+	__be32 *p;
+
+	p = reserve_space(xdr, 28);
+	p = xdr_encode_hyper(p, lowner->clientid);
+	*p++ = cpu_to_be32(16);
+	p = xdr_encode_opaque_fixed(p, "lock id:", 8);
+	xdr_encode_hyper(p, lowner->id);
+}
+
 /*
  * opcode,type,reclaim,offset,length,new_lock_owner = 32
  * open_seqid,open_stateid,lock_seqid,lock_owner.clientid, lock_owner.id = 40
@@ -1058,14 +1083,11 @@ static void encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args
 	p = xdr_encode_hyper(p, nfs4_lock_length(args->fl));
 	*p = cpu_to_be32(args->new_lock_owner);
 	if (args->new_lock_owner){
-		p = reserve_space(xdr, 4+NFS4_STATEID_SIZE+32);
+		p = reserve_space(xdr, 4+NFS4_STATEID_SIZE+4);
 		*p++ = cpu_to_be32(args->open_seqid->sequence->counter);
 		p = xdr_encode_opaque_fixed(p, args->open_stateid->data, NFS4_STATEID_SIZE);
 		*p++ = cpu_to_be32(args->lock_seqid->sequence->counter);
-		p = xdr_encode_hyper(p, args->lock_owner.clientid);
-		*p++ = cpu_to_be32(16);
-		p = xdr_encode_opaque_fixed(p, "lock id:", 8);
-		xdr_encode_hyper(p, args->lock_owner.id);
+		encode_lockowner(xdr, &args->lock_owner);
 	}
 	else {
 		p = reserve_space(xdr, NFS4_STATEID_SIZE+4);
@@ -1080,15 +1102,12 @@ static void encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *ar
 {
 	__be32 *p;
 
-	p = reserve_space(xdr, 52);
+	p = reserve_space(xdr, 24);
 	*p++ = cpu_to_be32(OP_LOCKT);
 	*p++ = cpu_to_be32(nfs4_lock_type(args->fl, 0));
 	p = xdr_encode_hyper(p, args->fl->fl_start);
 	p = xdr_encode_hyper(p, nfs4_lock_length(args->fl));
-	p = xdr_encode_hyper(p, args->lock_owner.clientid);
-	*p++ = cpu_to_be32(16);
-	p = xdr_encode_opaque_fixed(p, "lock id:", 8);
-	xdr_encode_hyper(p, args->lock_owner.id);
+	encode_lockowner(xdr, &args->lock_owner);
 	hdr->nops++;
 	hdr->replen += decode_lockt_maxsz;
 }
@@ -1108,6 +1127,17 @@ static void encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *ar
 	hdr->replen += decode_locku_maxsz;
 }
 
+static void encode_release_lockowner(struct xdr_stream *xdr, const struct nfs_lowner *lowner, struct compound_hdr *hdr)
+{
+	__be32 *p;
+
+	p = reserve_space(xdr, 4);
+	*p = cpu_to_be32(OP_RELEASE_LOCKOWNER);
+	encode_lockowner(xdr, lowner);
+	hdr->nops++;
+	hdr->replen += decode_release_lockowner_maxsz;
+}
+
 static void encode_lookup(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr)
 {
 	int len = name->len;
@@ -1172,7 +1202,7 @@ static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_op
 		break;
 	default:
 		clp = arg->server->nfs_client;
-		if (clp->cl_minorversion > 0) {
+		if (clp->cl_mvops->minor_version > 0) {
 			if (nfs4_has_persistent_session(clp)) {
 				*p = cpu_to_be32(NFS4_CREATE_GUARDED);
 				encode_attrs(xdr, arg->u.attrs, arg->server);
@@ -1324,14 +1354,14 @@ static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr)
 	hdr->replen += decode_putrootfh_maxsz;
 }
 
-static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx)
+static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx, const struct nfs_lock_context *l_ctx)
 {
 	nfs4_stateid stateid;
 	__be32 *p;
 
 	p = reserve_space(xdr, NFS4_STATEID_SIZE);
 	if (ctx->state != NULL) {
-		nfs4_copy_stateid(&stateid, ctx->state, ctx->lockowner);
+		nfs4_copy_stateid(&stateid, ctx->state, l_ctx->lockowner, l_ctx->pid);
 		xdr_encode_opaque_fixed(p, stateid.data, NFS4_STATEID_SIZE);
 	} else
 		xdr_encode_opaque_fixed(p, zero_stateid.data, NFS4_STATEID_SIZE);
@@ -1344,7 +1374,7 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args,
 	p = reserve_space(xdr, 4);
 	*p = cpu_to_be32(OP_READ);
 
-	encode_stateid(xdr, args->context);
+	encode_stateid(xdr, args->context, args->lock_context);
 
 	p = reserve_space(xdr, 12);
 	p = xdr_encode_hyper(p, args->offset);
@@ -1523,7 +1553,7 @@ static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *arg
 	p = reserve_space(xdr, 4);
 	*p = cpu_to_be32(OP_WRITE);
 
-	encode_stateid(xdr, args->context);
+	encode_stateid(xdr, args->context, args->lock_context);
 
 	p = reserve_space(xdr, 16);
 	p = xdr_encode_hyper(p, args->offset);
@@ -1704,7 +1734,7 @@ static u32 nfs4_xdr_minorversion(const struct nfs4_sequence_args *args)
 {
 #if defined(CONFIG_NFS_V4_1)
 	if (args->sa_session)
-		return args->sa_session->clp->cl_minorversion;
+		return args->sa_session->clp->cl_mvops->minor_version;
 #endif /* CONFIG_NFS_V4_1 */
 	return 0;
 }
@@ -2048,6 +2078,20 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_
 	return 0;
 }
 
+static int nfs4_xdr_enc_release_lockowner(struct rpc_rqst *req, __be32 *p, struct nfs_release_lockowner_args *args)
+{
+	struct xdr_stream xdr;
+	struct compound_hdr hdr = {
+		.minorversion = 0,
+	};
+
+	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
+	encode_compound_hdr(&xdr, req, &hdr);
+	encode_release_lockowner(&xdr, &args->lock_owner, &hdr);
+	encode_nops(&hdr);
+	return 0;
+}
+
 /*
  * Encode a READLINK request
  */
@@ -2395,7 +2439,7 @@ static int nfs4_xdr_enc_exchange_id(struct rpc_rqst *req, uint32_t *p,
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr = {
-		.minorversion = args->client->cl_minorversion,
+		.minorversion = args->client->cl_mvops->minor_version,
 	};
 
 	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
@@ -2413,7 +2457,7 @@ static int nfs4_xdr_enc_create_session(struct rpc_rqst *req, uint32_t *p,
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr = {
-		.minorversion = args->client->cl_minorversion,
+		.minorversion = args->client->cl_mvops->minor_version,
 	};
 
 	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
@@ -2431,7 +2475,7 @@ static int nfs4_xdr_enc_destroy_session(struct rpc_rqst *req, uint32_t *p,
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr = {
-		.minorversion = session->clp->cl_minorversion,
+		.minorversion = session->clp->cl_mvops->minor_version,
 	};
 
 	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
@@ -3973,6 +4017,11 @@ static int decode_locku(struct xdr_stream *xdr, struct nfs_locku_res *res)
 	return status;
 }
 
+static int decode_release_lockowner(struct xdr_stream *xdr)
+{
+	return decode_op_hdr(xdr, OP_RELEASE_LOCKOWNER);
+}
+
 static int decode_lookup(struct xdr_stream *xdr)
 {
 	return decode_op_hdr(xdr, OP_LOOKUP);
@@ -5259,6 +5308,19 @@ out:
 	return status;
 }
 
+static int nfs4_xdr_dec_release_lockowner(struct rpc_rqst *rqstp, __be32 *p, void *dummy)
+{
+	struct xdr_stream xdr;
+	struct compound_hdr hdr;
+	int status;
+
+	xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p);
+	status = decode_compound_hdr(&xdr, &hdr);
+	if (!status)
+		status = decode_release_lockowner(&xdr);
+	return status;
+}
+
 /*
  * Decode READLINK response
  */
@@ -5866,6 +5928,7 @@ struct rpc_procinfo	nfs4_procedures[] = {
   PROC(GETACL,		enc_getacl,	dec_getacl),
   PROC(SETACL,		enc_setacl,	dec_setacl),
   PROC(FS_LOCATIONS,	enc_fs_locations, dec_fs_locations),
+  PROC(RELEASE_LOCKOWNER, enc_release_lockowner, dec_release_lockowner),
 #if defined(CONFIG_NFS_V4_1)
   PROC(EXCHANGE_ID,	enc_exchange_id,	dec_exchange_id),
   PROC(CREATE_SESSION,	enc_create_session,	dec_create_session),
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index a3654e57b58..919490232e1 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -79,6 +79,7 @@ nfs_create_request(struct nfs_open_context *ctx, struct inode *inode,
 	req->wb_pgbase	= offset;
 	req->wb_bytes   = count;
 	req->wb_context = get_nfs_open_context(ctx);
+	req->wb_lock_context = nfs_get_lock_context(ctx);
 	kref_init(&req->wb_kref);
 	return req;
 }
@@ -141,11 +142,16 @@ void nfs_clear_request(struct nfs_page *req)
 {
 	struct page *page = req->wb_page;
 	struct nfs_open_context *ctx = req->wb_context;
+	struct nfs_lock_context *l_ctx = req->wb_lock_context;
 
 	if (page != NULL) {
 		page_cache_release(page);
 		req->wb_page = NULL;
 	}
+	if (l_ctx != NULL) {
+		nfs_put_lock_context(l_ctx);
+		req->wb_lock_context = NULL;
+	}
 	if (ctx != NULL) {
 		put_nfs_open_context(ctx);
 		req->wb_context = NULL;
@@ -235,7 +241,7 @@ static int nfs_can_coalesce_requests(struct nfs_page *prev,
 {
 	if (req->wb_context->cred != prev->wb_context->cred)
 		return 0;
-	if (req->wb_context->lockowner != prev->wb_context->lockowner)
+	if (req->wb_lock_context->lockowner != prev->wb_lock_context->lockowner)
 		return 0;
 	if (req->wb_context->state != prev->wb_context->state)
 		return 0;
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 6e2b06e6ca7..87adc274424 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -190,6 +190,7 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
 	data->args.pages  = data->pagevec;
 	data->args.count  = count;
 	data->args.context = get_nfs_open_context(req->wb_context);
+	data->args.lock_context = req->wb_lock_context;
 
 	data->res.fattr   = &data->fattr;
 	data->res.count   = count;
@@ -410,7 +411,7 @@ void nfs_read_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs_read_data *data = calldata;
 
-	if (nfs4_setup_sequence(NFS_SERVER(data->inode)->nfs_client,
+	if (nfs4_setup_sequence(NFS_SERVER(data->inode),
 				&data->args.seq_args, &data->res.seq_res,
 				0, task))
 		return;
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index f9df16de4a5..f1ae39f6cb0 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -546,6 +546,9 @@ static void nfs_show_mountd_options(struct seq_file *m, struct nfs_server *nfss,
 {
 	struct sockaddr *sap = (struct sockaddr *)&nfss->mountd_address;
 
+	if (nfss->flags & NFS_MOUNT_LEGACY_INTERFACE)
+		return;
+
 	switch (sap->sa_family) {
 	case AF_INET: {
 		struct sockaddr_in *sin = (struct sockaddr_in *)sap;
@@ -1780,6 +1783,7 @@ static int nfs_validate_mount_data(void *options,
 		 * can deal with.
 		 */
 		args->flags		= data->flags & NFS_MOUNT_FLAGMASK;
+		args->flags		|= NFS_MOUNT_LEGACY_INTERFACE;
 		args->rsize		= data->rsize;
 		args->wsize		= data->wsize;
 		args->timeo		= data->timeo;
diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c
index a2242af6a17..2f84adaad42 100644
--- a/fs/nfs/unlink.c
+++ b/fs/nfs/unlink.c
@@ -110,7 +110,7 @@ void nfs_unlink_prepare(struct rpc_task *task, void *calldata)
 	struct nfs_unlinkdata *data = calldata;
 	struct nfs_server *server = NFS_SERVER(data->dir);
 
-	if (nfs4_setup_sequence(server->nfs_client, &data->args.seq_args,
+	if (nfs4_setup_sequence(server, &data->args.seq_args,
 				&data->res.seq_res, 1, task))
 		return;
 	rpc_call_start(task);
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 9f81bdd91c5..874972d9427 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -700,7 +700,9 @@ int nfs_flush_incompatible(struct file *file, struct page *page)
 		req = nfs_page_find_request(page);
 		if (req == NULL)
 			return 0;
-		do_flush = req->wb_page != page || req->wb_context != ctx;
+		do_flush = req->wb_page != page || req->wb_context != ctx ||
+			req->wb_lock_context->lockowner != current->files ||
+			req->wb_lock_context->pid != current->tgid;
 		nfs_release_request(req);
 		if (!do_flush)
 			return 0;
@@ -824,6 +826,7 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 	data->args.pages  = data->pagevec;
 	data->args.count  = count;
 	data->args.context = get_nfs_open_context(req->wb_context);
+	data->args.lock_context = req->wb_lock_context;
 	data->args.stable  = NFS_UNSTABLE;
 	if (how & FLUSH_STABLE) {
 		data->args.stable = NFS_DATA_SYNC;
@@ -1047,9 +1050,9 @@ out:
 void nfs_write_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs_write_data *data = calldata;
-	struct nfs_client *clp = (NFS_SERVER(data->inode))->nfs_client;
 
-	if (nfs4_setup_sequence(clp, &data->args.seq_args,
+	if (nfs4_setup_sequence(NFS_SERVER(data->inode),
+				&data->args.seq_args,
 				&data->res.seq_res, 1, task))
 		return;
 	rpc_call_start(task);
diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
index 3d68f45a37b..5b7e3021e06 100644
--- a/fs/nfsd/nfs3proc.c
+++ b/fs/nfsd/nfs3proc.c
@@ -168,7 +168,7 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp,
 	svc_reserve_auth(rqstp, ((1 + NFS3_POST_OP_ATTR_WORDS + 3)<<2) + resp->count +4);
 
 	fh_copy(&resp->fh, &argp->fh);
-	nfserr = nfsd_read(rqstp, &resp->fh, NULL,
+	nfserr = nfsd_read(rqstp, &resp->fh,
 				  argp->offset,
 			   	  rqstp->rq_vec, argp->vlen,
 				  &resp->count);
@@ -271,7 +271,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp, struct nfsd3_createargs *argp,
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				    &argp->attrs, S_IFDIR, 0, &resp->fh);
-
+	fh_unlock(&resp->dirfh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -327,7 +327,7 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp, struct nfsd3_mknodargs *argp,
 	type = nfs3_ftypes[argp->ftype];
 	nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				    &argp->attrs, type, rdev, &resp->fh);
-
+	fh_unlock(&resp->dirfh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -348,6 +348,7 @@ nfsd3_proc_remove(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp,
 	/* Unlink. -S_IFDIR means file must not be a directory */
 	fh_copy(&resp->fh, &argp->fh);
 	nfserr = nfsd_unlink(rqstp, &resp->fh, -S_IFDIR, argp->name, argp->len);
+	fh_unlock(&resp->fh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -367,6 +368,7 @@ nfsd3_proc_rmdir(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp,
 
 	fh_copy(&resp->fh, &argp->fh);
 	nfserr = nfsd_unlink(rqstp, &resp->fh, S_IFDIR, argp->name, argp->len);
+	fh_unlock(&resp->fh);
 	RETURN_STATUS(nfserr);
 }
 
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index eb78e7e2207..988cbb3a19b 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -143,8 +143,6 @@ struct nfs4_cb_compound_hdr {
 	u32		minorversion;
 	/* res */
 	int		status;
-	u32		taglen;
-	char		*tag;
 };
 
 static struct {
@@ -205,6 +203,16 @@ nfs_cb_stat_to_errno(int stat)
  */
 
 static void
+encode_stateid(struct xdr_stream *xdr, stateid_t *sid)
+{
+	__be32 *p;
+
+	RESERVE_SPACE(sizeof(stateid_t));
+	WRITE32(sid->si_generation);
+	WRITEMEM(&sid->si_opaque, sizeof(stateid_opaque_t));
+}
+
+static void
 encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr)
 {
 	__be32 * p;
@@ -229,10 +237,10 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp,
 	__be32 *p;
 	int len = dp->dl_fh.fh_size;
 
-	RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len);
+	RESERVE_SPACE(4);
 	WRITE32(OP_CB_RECALL);
-	WRITE32(dp->dl_stateid.si_generation);
-	WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t));
+	encode_stateid(xdr, &dp->dl_stateid);
+	RESERVE_SPACE(8 + (XDR_QUADLEN(len) << 2));
 	WRITE32(0); /* truncate optimization not implemented */
 	WRITE32(len);
 	WRITEMEM(&dp->dl_fh.fh_base, len);
@@ -293,13 +301,14 @@ nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p,
 static int
 decode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr){
         __be32 *p;
+	u32 taglen;
 
         READ_BUF(8);
         READ32(hdr->status);
-        READ32(hdr->taglen);
-        READ_BUF(hdr->taglen + 4);
-        hdr->tag = (char *)p;
-        p += XDR_QUADLEN(hdr->taglen);
+	/* We've got no use for the tag; ignore it: */
+        READ32(taglen);
+        READ_BUF(taglen + 4);
+        p += XDR_QUADLEN(taglen);
         READ32(hdr->nops);
         return 0;
 }
@@ -667,28 +676,28 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 	}
 
 	switch (task->tk_status) {
-	case -EIO:
+	case 0:
+		return;
+	case -EBADHANDLE:
+	case -NFS4ERR_BAD_STATEID:
+		/* Race: client probably got cb_recall
+		 * before open reply granting delegation */
+		break;
+	default:
 		/* Network partition? */
 		atomic_set(&clp->cl_cb_set, 0);
 		warn_no_callback_path(clp, task->tk_status);
 		if (current_rpc_client != task->tk_client) {
 			/* queue a callback on the new connection: */
+			atomic_inc(&dp->dl_count);
 			nfsd4_cb_recall(dp);
 			return;
 		}
-	case -EBADHANDLE:
-	case -NFS4ERR_BAD_STATEID:
-		/* Race: client probably got cb_recall
-		 * before open reply granting delegation */
-		break;
-	default:
-		/* success, or error we can't handle */
-		return;
 	}
 	if (dp->dl_retries--) {
 		rpc_delay(task, 2*HZ);
 		task->tk_status = 0;
-		rpc_restart_call(task);
+		rpc_restart_call_prepare(task);
 		return;
 	} else {
 		atomic_set(&clp->cl_cb_set, 0);
@@ -752,18 +761,16 @@ static void _nfsd4_cb_recall(struct nfs4_delegation *dp)
 		.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL],
 		.rpc_cred = callback_cred
 	};
-	int status;
 
-	if (clnt == NULL)
+	if (clnt == NULL) {
+		nfs4_put_delegation(dp);
 		return; /* Client is shutting down; give up. */
+	}
 
 	args->args_op = dp;
 	msg.rpc_argp = args;
 	dp->dl_retries = 1;
-	status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT,
-				&nfsd4_cb_recall_ops, dp);
-	if (status)
-		nfs4_put_delegation(dp);
+	rpc_call_async(clnt, &msg, RPC_TASK_SOFT, &nfsd4_cb_recall_ops, dp);
 }
 
 void nfsd4_do_callback_rpc(struct work_struct *w)
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 4a273475877..2e7357104cf 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -51,7 +51,6 @@ static time_t boot_time;
 static u32 current_ownerid = 1;
 static u32 current_fileid = 1;
 static u32 current_delegid = 1;
-static u32 nfs4_init;
 static stateid_t zerostateid;             /* bits all 0 */
 static stateid_t onestateid;              /* bits all 1 */
 static u64 current_sessionid = 1;
@@ -163,6 +162,46 @@ static struct list_head	ownerstr_hashtbl[OWNER_HASH_SIZE];
 static struct list_head file_hashtbl[FILE_HASH_SIZE];
 static struct list_head stateid_hashtbl[STATEID_HASH_SIZE];
 
+static void __nfs4_file_get_access(struct nfs4_file *fp, int oflag)
+{
+	BUG_ON(!(fp->fi_fds[oflag] || fp->fi_fds[O_RDWR]));
+	atomic_inc(&fp->fi_access[oflag]);
+}
+
+static void nfs4_file_get_access(struct nfs4_file *fp, int oflag)
+{
+	if (oflag == O_RDWR) {
+		__nfs4_file_get_access(fp, O_RDONLY);
+		__nfs4_file_get_access(fp, O_WRONLY);
+	} else
+		__nfs4_file_get_access(fp, oflag);
+}
+
+static void nfs4_file_put_fd(struct nfs4_file *fp, int oflag)
+{
+	if (fp->fi_fds[oflag]) {
+		fput(fp->fi_fds[oflag]);
+		fp->fi_fds[oflag] = NULL;
+	}
+}
+
+static void __nfs4_file_put_access(struct nfs4_file *fp, int oflag)
+{
+	if (atomic_dec_and_test(&fp->fi_access[oflag])) {
+		nfs4_file_put_fd(fp, O_RDWR);
+		nfs4_file_put_fd(fp, oflag);
+	}
+}
+
+static void nfs4_file_put_access(struct nfs4_file *fp, int oflag)
+{
+	if (oflag == O_RDWR) {
+		__nfs4_file_put_access(fp, O_RDONLY);
+		__nfs4_file_put_access(fp, O_WRONLY);
+	} else
+		__nfs4_file_put_access(fp, oflag);
+}
+
 static struct nfs4_delegation *
 alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_fh *current_fh, u32 type)
 {
@@ -171,6 +210,13 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
 	struct nfs4_cb_conn *cb = &stp->st_stateowner->so_client->cl_cb_conn;
 
 	dprintk("NFSD alloc_init_deleg\n");
+	/*
+	 * Major work on the lease subsystem (for example, to support
+	 * calbacks on stat) will be required before we can support
+	 * write delegations properly.
+	 */
+	if (type != NFS4_OPEN_DELEGATE_READ)
+		return NULL;
 	if (fp->fi_had_conflict)
 		return NULL;
 	if (num_delegations > max_delegations)
@@ -185,9 +231,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
 	dp->dl_client = clp;
 	get_nfs4_file(fp);
 	dp->dl_file = fp;
+	nfs4_file_get_access(fp, O_RDONLY);
 	dp->dl_flock = NULL;
-	get_file(stp->st_vfs_file);
-	dp->dl_vfs_file = stp->st_vfs_file;
 	dp->dl_type = type;
 	dp->dl_ident = cb->cb_ident;
 	dp->dl_stateid.si_boot = boot_time;
@@ -222,15 +267,12 @@ nfs4_put_delegation(struct nfs4_delegation *dp)
 static void
 nfs4_close_delegation(struct nfs4_delegation *dp)
 {
-	struct file *filp = dp->dl_vfs_file;
+	struct file *filp = find_readable_file(dp->dl_file);
 
 	dprintk("NFSD: close_delegation dp %p\n",dp);
-	dp->dl_vfs_file = NULL;
-	/* The following nfsd_close may not actually close the file,
-	 * but we want to remove the lease in any case. */
 	if (dp->dl_flock)
 		vfs_setlease(filp, F_UNLCK, &dp->dl_flock);
-	nfsd_close(filp);
+	nfs4_file_put_access(dp->dl_file, O_RDONLY);
 }
 
 /* Called under the state lock. */
@@ -302,8 +344,12 @@ static void free_generic_stateid(struct nfs4_stateid *stp)
 
 static void release_lock_stateid(struct nfs4_stateid *stp)
 {
+	struct file *file;
+
 	unhash_generic_stateid(stp);
-	locks_remove_posix(stp->st_vfs_file, (fl_owner_t)stp->st_stateowner);
+	file = find_any_file(stp->st_file);
+	if (file)
+		locks_remove_posix(file, (fl_owner_t)stp->st_stateowner);
 	free_generic_stateid(stp);
 }
 
@@ -341,11 +387,85 @@ release_stateid_lockowners(struct nfs4_stateid *open_stp)
 	}
 }
 
+/*
+ * We store the NONE, READ, WRITE, and BOTH bits separately in the
+ * st_{access,deny}_bmap field of the stateid, in order to track not
+ * only what share bits are currently in force, but also what
+ * combinations of share bits previous opens have used.  This allows us
+ * to enforce the recommendation of rfc 3530 14.2.19 that the server
+ * return an error if the client attempt to downgrade to a combination
+ * of share bits not explicable by closing some of its previous opens.
+ *
+ * XXX: This enforcement is actually incomplete, since we don't keep
+ * track of access/deny bit combinations; so, e.g., we allow:
+ *
+ *	OPEN allow read, deny write
+ *	OPEN allow both, deny none
+ *	DOWNGRADE allow read, deny none
+ *
+ * which we should reject.
+ */
+static void
+set_access(unsigned int *access, unsigned long bmap) {
+	int i;
+
+	*access = 0;
+	for (i = 1; i < 4; i++) {
+		if (test_bit(i, &bmap))
+			*access |= i;
+	}
+}
+
+static void
+set_deny(unsigned int *deny, unsigned long bmap) {
+	int i;
+
+	*deny = 0;
+	for (i = 0; i < 4; i++) {
+		if (test_bit(i, &bmap))
+			*deny |= i ;
+	}
+}
+
+static int
+test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) {
+	unsigned int access, deny;
+
+	set_access(&access, stp->st_access_bmap);
+	set_deny(&deny, stp->st_deny_bmap);
+	if ((access & open->op_share_deny) || (deny & open->op_share_access))
+		return 0;
+	return 1;
+}
+
+static int nfs4_access_to_omode(u32 access)
+{
+	switch (access) {
+	case NFS4_SHARE_ACCESS_READ:
+		return O_RDONLY;
+	case NFS4_SHARE_ACCESS_WRITE:
+		return O_WRONLY;
+	case NFS4_SHARE_ACCESS_BOTH:
+		return O_RDWR;
+	}
+	BUG();
+}
+
+static int nfs4_access_bmap_to_omode(struct nfs4_stateid *stp)
+{
+	unsigned int access;
+
+	set_access(&access, stp->st_access_bmap);
+	return nfs4_access_to_omode(access);
+}
+
 static void release_open_stateid(struct nfs4_stateid *stp)
 {
+	int oflag = nfs4_access_bmap_to_omode(stp);
+
 	unhash_generic_stateid(stp);
 	release_stateid_lockowners(stp);
-	nfsd_close(stp->st_vfs_file);
+	nfs4_file_put_access(stp->st_file, oflag);
 	free_generic_stateid(stp);
 }
 
@@ -457,7 +577,7 @@ static int set_forechannel_drc_size(struct nfsd4_channel_attrs *fchan)
 	spin_unlock(&nfsd_drc_lock);
 
 	if (fchan->maxreqs == 0)
-		return nfserr_serverfault;
+		return nfserr_jukebox;
 
 	fchan->maxresp_cached = size + NFSD_MIN_HDR_SEQ_SZ;
 	return 0;
@@ -542,7 +662,7 @@ alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp,
 	BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot)
 		     + sizeof(struct nfsd4_session) > PAGE_SIZE);
 
-	status = nfserr_serverfault;
+	status = nfserr_jukebox;
 	/* allocate struct nfsd4_session and slot table pointers in one piece */
 	slotsize = tmp.se_fchannel.maxreqs * sizeof(struct nfsd4_slot *);
 	new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL);
@@ -591,10 +711,8 @@ find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid)
 
 	dump_sessionid(__func__, sessionid);
 	idx = hash_sessionid(sessionid);
-	dprintk("%s: idx is %d\n", __func__, idx);
 	/* Search in the appropriate list */
 	list_for_each_entry(elem, &sessionid_hashtbl[idx], se_hash) {
-		dump_sessionid("list traversal", &elem->se_sessionid);
 		if (!memcmp(elem->se_sessionid.data, sessionid->data,
 			    NFS4_MAX_SESSIONID_LEN)) {
 			return elem;
@@ -714,7 +832,6 @@ release_session_client(struct nfsd4_session *session)
 	} else
 		renew_client_locked(clp);
 	spin_unlock(&client_lock);
-	nfsd4_put_session(session);
 }
 
 /* must be called under the client_lock */
@@ -1220,7 +1337,7 @@ out_new:
 	/* Normal case */
 	new = create_client(exid->clname, dname, rqstp, &verf);
 	if (new == NULL) {
-		status = nfserr_serverfault;
+		status = nfserr_jukebox;
 		goto out;
 	}
 
@@ -1760,6 +1877,8 @@ alloc_init_file(struct inode *ino)
 		fp->fi_inode = igrab(ino);
 		fp->fi_id = current_fileid++;
 		fp->fi_had_conflict = false;
+		memset(fp->fi_fds, 0, sizeof(fp->fi_fds));
+		memset(fp->fi_access, 0, sizeof(fp->fi_access));
 		spin_lock(&recall_lock);
 		list_add(&fp->fi_hash, &file_hashtbl[hashval]);
 		spin_unlock(&recall_lock);
@@ -1971,57 +2090,6 @@ static inline int deny_valid(u32 x)
 }
 
 /*
- * We store the NONE, READ, WRITE, and BOTH bits separately in the
- * st_{access,deny}_bmap field of the stateid, in order to track not
- * only what share bits are currently in force, but also what
- * combinations of share bits previous opens have used.  This allows us
- * to enforce the recommendation of rfc 3530 14.2.19 that the server
- * return an error if the client attempt to downgrade to a combination
- * of share bits not explicable by closing some of its previous opens.
- *
- * XXX: This enforcement is actually incomplete, since we don't keep
- * track of access/deny bit combinations; so, e.g., we allow:
- *
- *	OPEN allow read, deny write
- *	OPEN allow both, deny none
- *	DOWNGRADE allow read, deny none
- *
- * which we should reject.
- */
-static void
-set_access(unsigned int *access, unsigned long bmap) {
-	int i;
-
-	*access = 0;
-	for (i = 1; i < 4; i++) {
-		if (test_bit(i, &bmap))
-			*access |= i;
-	}
-}
-
-static void
-set_deny(unsigned int *deny, unsigned long bmap) {
-	int i;
-
-	*deny = 0;
-	for (i = 0; i < 4; i++) {
-		if (test_bit(i, &bmap))
-			*deny |= i ;
-	}
-}
-
-static int
-test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) {
-	unsigned int access, deny;
-
-	set_access(&access, stp->st_access_bmap);
-	set_deny(&deny, stp->st_deny_bmap);
-	if ((access & open->op_share_deny) || (deny & open->op_share_access))
-		return 0;
-	return 1;
-}
-
-/*
  * Called to check deny when READ with all zero stateid or
  * WRITE with all zero or all one stateid
  */
@@ -2052,14 +2120,12 @@ out:
 }
 
 static inline void
-nfs4_file_downgrade(struct file *filp, unsigned int share_access)
+nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access)
 {
-	if (share_access & NFS4_SHARE_ACCESS_WRITE) {
-		drop_file_write_access(filp);
-		spin_lock(&filp->f_lock);
-		filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE;
-		spin_unlock(&filp->f_lock);
-	}
+	if (share_access & NFS4_SHARE_ACCESS_WRITE)
+		nfs4_file_put_access(fp, O_WRONLY);
+	if (share_access & NFS4_SHARE_ACCESS_READ)
+		nfs4_file_put_access(fp, O_RDONLY);
 }
 
 /*
@@ -2255,6 +2321,13 @@ find_delegation_file(struct nfs4_file *fp, stateid_t *stid)
 	return NULL;
 }
 
+int share_access_to_flags(u32 share_access)
+{
+	share_access &= ~NFS4_SHARE_WANT_MASK;
+
+	return share_access == NFS4_SHARE_ACCESS_READ ? RD_STATE : WR_STATE;
+}
+
 static __be32
 nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open,
 		struct nfs4_delegation **dp)
@@ -2265,8 +2338,7 @@ nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open,
 	*dp = find_delegation_file(fp, &open->op_delegate_stateid);
 	if (*dp == NULL)
 		goto out;
-	flags = open->op_share_access == NFS4_SHARE_ACCESS_READ ?
-						RD_STATE : WR_STATE;
+	flags = share_access_to_flags(open->op_share_access);
 	status = nfs4_check_delegmode(*dp, flags);
 	if (status)
 		*dp = NULL;
@@ -2308,30 +2380,53 @@ nfs4_alloc_stateid(void)
 	return kmem_cache_alloc(stateid_slab, GFP_KERNEL);
 }
 
+static inline int nfs4_access_to_access(u32 nfs4_access)
+{
+	int flags = 0;
+
+	if (nfs4_access & NFS4_SHARE_ACCESS_READ)
+		flags |= NFSD_MAY_READ;
+	if (nfs4_access & NFS4_SHARE_ACCESS_WRITE)
+		flags |= NFSD_MAY_WRITE;
+	return flags;
+}
+
+static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file
+*fp, struct svc_fh *cur_fh, u32 nfs4_access)
+{
+	__be32 status;
+	int oflag = nfs4_access_to_omode(nfs4_access);
+	int access = nfs4_access_to_access(nfs4_access);
+
+	if (!fp->fi_fds[oflag]) {
+		status = nfsd_open(rqstp, cur_fh, S_IFREG, access,
+			&fp->fi_fds[oflag]);
+		if (status == nfserr_dropit)
+			status = nfserr_jukebox;
+		if (status)
+			return status;
+	}
+	nfs4_file_get_access(fp, oflag);
+
+	return nfs_ok;
+}
+
 static __be32
 nfs4_new_open(struct svc_rqst *rqstp, struct nfs4_stateid **stpp,
-		struct nfs4_delegation *dp,
-		struct svc_fh *cur_fh, int flags)
+		struct nfs4_file *fp, struct svc_fh *cur_fh,
+		struct nfsd4_open *open)
 {
 	struct nfs4_stateid *stp;
+	__be32 status;
 
 	stp = nfs4_alloc_stateid();
 	if (stp == NULL)
 		return nfserr_resource;
 
-	if (dp) {
-		get_file(dp->dl_vfs_file);
-		stp->st_vfs_file = dp->dl_vfs_file;
-	} else {
-		__be32 status;
-		status = nfsd_open(rqstp, cur_fh, S_IFREG, flags,
-				&stp->st_vfs_file);
-		if (status) {
-			if (status == nfserr_dropit)
-				status = nfserr_jukebox;
-			kmem_cache_free(stateid_slab, stp);
-			return status;
-		}
+	status = nfs4_get_vfs_file(rqstp, fp, cur_fh, open->op_share_access);
+	if (status) {
+		kmem_cache_free(stateid_slab, stp);
+		return status;
 	}
 	*stpp = stp;
 	return 0;
@@ -2353,35 +2448,30 @@ nfsd4_truncate(struct svc_rqst *rqstp, struct svc_fh *fh,
 }
 
 static __be32
-nfs4_upgrade_open(struct svc_rqst *rqstp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open)
+nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open)
 {
-	struct file *filp = stp->st_vfs_file;
-	struct inode *inode = filp->f_path.dentry->d_inode;
-	unsigned int share_access, new_writer;
+	u32 op_share_access, new_access;
 	__be32 status;
 
-	set_access(&share_access, stp->st_access_bmap);
-	new_writer = (~share_access) & open->op_share_access
-			& NFS4_SHARE_ACCESS_WRITE;
-
-	if (new_writer) {
-		int err = get_write_access(inode);
-		if (err)
-			return nfserrno(err);
-		err = mnt_want_write(cur_fh->fh_export->ex_path.mnt);
-		if (err)
-			return nfserrno(err);
-		file_take_write(filp);
+	set_access(&new_access, stp->st_access_bmap);
+	new_access = (~new_access) & open->op_share_access & ~NFS4_SHARE_WANT_MASK;
+
+	if (new_access) {
+		status = nfs4_get_vfs_file(rqstp, fp, cur_fh, new_access);
+		if (status)
+			return status;
 	}
 	status = nfsd4_truncate(rqstp, cur_fh, open);
 	if (status) {
-		if (new_writer)
-			put_write_access(inode);
+		if (new_access) {
+			int oflag = nfs4_access_to_omode(new_access);
+			nfs4_file_put_access(fp, oflag);
+		}
 		return status;
 	}
 	/* remember the open */
-	filp->f_mode |= open->op_share_access;
-	__set_bit(open->op_share_access, &stp->st_access_bmap);
+	op_share_access = open->op_share_access & ~NFS4_SHARE_WANT_MASK;
+	__set_bit(op_share_access, &stp->st_access_bmap);
 	__set_bit(open->op_share_deny, &stp->st_deny_bmap);
 
 	return nfs_ok;
@@ -2444,13 +2534,14 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta
 	fl.fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK;
 	fl.fl_end = OFFSET_MAX;
 	fl.fl_owner =  (fl_owner_t)dp;
-	fl.fl_file = stp->st_vfs_file;
+	fl.fl_file = find_readable_file(stp->st_file);
+	BUG_ON(!fl.fl_file);
 	fl.fl_pid = current->tgid;
 
 	/* vfs_setlease checks to see if delegation should be handed out.
 	 * the lock_manager callbacks fl_mylease and fl_change are used
 	 */
-	if ((status = vfs_setlease(stp->st_vfs_file, fl.fl_type, &flp))) {
+	if ((status = vfs_setlease(fl.fl_file, fl.fl_type, &flp))) {
 		dprintk("NFSD: setlease failed [%d], no delegation\n", status);
 		unhash_delegation(dp);
 		flag = NFS4_OPEN_DELEGATE_NONE;
@@ -2514,18 +2605,12 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	 */
 	if (stp) {
 		/* Stateid was found, this is an OPEN upgrade */
-		status = nfs4_upgrade_open(rqstp, current_fh, stp, open);
+		status = nfs4_upgrade_open(rqstp, fp, current_fh, stp, open);
 		if (status)
 			goto out;
 		update_stateid(&stp->st_stateid);
 	} else {
-		/* Stateid was not found, this is a new OPEN */
-		int flags = 0;
-		if (open->op_share_access & NFS4_SHARE_ACCESS_READ)
-			flags |= NFSD_MAY_READ;
-		if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE)
-			flags |= NFSD_MAY_WRITE;
-		status = nfs4_new_open(rqstp, &stp, dp, current_fh, flags);
+		status = nfs4_new_open(rqstp, &stp, fp, current_fh, open);
 		if (status)
 			goto out;
 		init_stateid(stp, fp, open);
@@ -2727,7 +2812,7 @@ search_close_lru(u32 st_id, int flags)
 static inline int
 nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp)
 {
-	return fhp->fh_dentry->d_inode != stp->st_vfs_file->f_path.dentry->d_inode;
+	return fhp->fh_dentry->d_inode != stp->st_file->fi_inode;
 }
 
 static int
@@ -2760,6 +2845,9 @@ __be32 nfs4_check_openmode(struct nfs4_stateid *stp, int flags)
 {
         __be32 status = nfserr_openmode;
 
+	/* For lock stateid's, we test the parent open, not the lock: */
+	if (stp->st_openstp)
+		stp = stp->st_openstp;
 	if ((flags & WR_STATE) && (!access_permit_write(stp->st_access_bmap)))
                 goto out;
 	if ((flags & RD_STATE) && (!access_permit_read(stp->st_access_bmap)))
@@ -2872,7 +2960,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 			goto out;
 		renew_client(dp->dl_client);
 		if (filpp)
-			*filpp = dp->dl_vfs_file;
+			*filpp = find_readable_file(dp->dl_file);
+		BUG_ON(!*filpp);
 	} else { /* open or lock stateid */
 		stp = find_stateid(stateid, flags);
 		if (!stp)
@@ -2889,8 +2978,13 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 		if (status)
 			goto out;
 		renew_client(stp->st_stateowner->so_client);
-		if (filpp)
-			*filpp = stp->st_vfs_file;
+		if (filpp) {
+			if (flags & RD_STATE)
+				*filpp = find_readable_file(stp->st_file);
+			else
+				*filpp = find_writeable_file(stp->st_file);
+			BUG_ON(!*filpp); /* assured by check_openmode */
+		}
 	}
 	status = nfs_ok;
 out:
@@ -3126,8 +3220,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp,
 		goto out;
 	}
 	set_access(&share_access, stp->st_access_bmap);
-	nfs4_file_downgrade(stp->st_vfs_file,
-	                    share_access & ~od->od_share_access);
+	nfs4_file_downgrade(stp->st_file, share_access & ~od->od_share_access);
 
 	reset_union_bmap_access(od->od_share_access, &stp->st_access_bmap);
 	reset_union_bmap_deny(od->od_share_deny, &stp->st_deny_bmap);
@@ -3346,11 +3439,9 @@ static inline void
 nfs4_set_lock_denied(struct file_lock *fl, struct nfsd4_lock_denied *deny)
 {
 	struct nfs4_stateowner *sop;
-	unsigned int hval;
 
 	if (fl->fl_lmops == &nfsd_posix_mng_ops) {
 		sop = (struct nfs4_stateowner *) fl->fl_owner;
-		hval = lockownerid_hashval(sop->so_id);
 		kref_get(&sop->so_ref);
 		deny->ld_sop = sop;
 		deny->ld_clientid = sop->so_client->cl_clientid;
@@ -3446,8 +3537,6 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc
 	stp->st_stateid.si_stateownerid = sop->so_id;
 	stp->st_stateid.si_fileid = fp->fi_id;
 	stp->st_stateid.si_generation = 0;
-	stp->st_vfs_file = open_stp->st_vfs_file; /* FIXME refcount?? */
-	stp->st_access_bmap = open_stp->st_access_bmap;
 	stp->st_deny_bmap = open_stp->st_deny_bmap;
 	stp->st_openstp = open_stp;
 
@@ -3547,7 +3636,6 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		lock_sop = lock->lk_replay_owner;
 	}
 	/* lock->lk_replay_owner and lock_stp have been created or found */
-	filp = lock_stp->st_vfs_file;
 
 	status = nfserr_grace;
 	if (locks_in_grace() && !lock->lk_reclaim)
@@ -3560,11 +3648,13 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	switch (lock->lk_type) {
 		case NFS4_READ_LT:
 		case NFS4_READW_LT:
+			filp = find_readable_file(lock_stp->st_file);
 			file_lock.fl_type = F_RDLCK;
 			cmd = F_SETLK;
 		break;
 		case NFS4_WRITE_LT:
 		case NFS4_WRITEW_LT:
+			filp = find_writeable_file(lock_stp->st_file);
 			file_lock.fl_type = F_WRLCK;
 			cmd = F_SETLK;
 		break;
@@ -3572,6 +3662,10 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			status = nfserr_inval;
 		goto out;
 	}
+	if (!filp) {
+		status = nfserr_openmode;
+		goto out;
+	}
 	file_lock.fl_owner = (fl_owner_t)lock_sop;
 	file_lock.fl_pid = current->tgid;
 	file_lock.fl_file = filp;
@@ -3740,7 +3834,11 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 					&locku->lu_stateowner, &stp, NULL)))
 		goto out;
 
-	filp = stp->st_vfs_file;
+	filp = find_any_file(stp->st_file);
+	if (!filp) {
+		status = nfserr_lock_range;
+		goto out;
+	}
 	BUG_ON(!filp);
 	locks_init_lock(&file_lock);
 	file_lock.fl_type = F_UNLCK;
@@ -3787,10 +3885,10 @@ out_nfserr:
  * 	0: no locks held by lockowner
  */
 static int
-check_for_locks(struct file *filp, struct nfs4_stateowner *lowner)
+check_for_locks(struct nfs4_file *filp, struct nfs4_stateowner *lowner)
 {
 	struct file_lock **flpp;
-	struct inode *inode = filp->f_path.dentry->d_inode;
+	struct inode *inode = filp->fi_inode;
 	int status = 0;
 
 	lock_kernel();
@@ -3841,7 +3939,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 				continue;
 			list_for_each_entry(stp, &sop->so_stateids,
 					st_perstateowner) {
-				if (check_for_locks(stp->st_vfs_file, sop))
+				if (check_for_locks(stp->st_file, sop))
 					goto out;
 				/* Note: so_perclient unused for lockowners,
 				 * so it's OK to fool with here. */
@@ -4066,16 +4164,8 @@ out_free_laundry:
 int
 nfs4_state_start(void)
 {
-	int ret;
-
-	if (nfs4_init)
-		return 0;
 	nfsd4_load_reboot_recovery_data();
-	ret = __nfs4_state_start();
-	if (ret)
-		return ret;
-	nfs4_init = 1;
-	return 0;
+	return __nfs4_state_start();
 }
 
 static void
@@ -4110,7 +4200,6 @@ __nfs4_state_shutdown(void)
 	}
 
 	nfsd4_shutdown_recdir();
-	nfs4_init = 0;
 }
 
 void
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index ac17a708023..f8931acb05f 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -2630,7 +2630,7 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 	}
 	read->rd_vlen = v;
 
-	nfserr = nfsd_read(read->rd_rqstp, read->rd_fhp, read->rd_filp,
+	nfserr = nfsd_read_file(read->rd_rqstp, read->rd_fhp, read->rd_filp,
 			read->rd_offset, resp->rqstp->rq_vec, read->rd_vlen,
 			&maxcount);
 
@@ -3325,6 +3325,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo
 		}
 		/* Renew the clientid on success and on replay */
 		release_session_client(cs->session);
+		nfsd4_put_session(cs->session);
 	}
 	return 1;
 }
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 508941c23af..b53b1d042f1 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -949,15 +949,12 @@ static ssize_t __write_ports_addfd(char *buf)
 	if (err != 0)
 		return err;
 
-	err = lockd_up();
-	if (err != 0)
-		goto out;
-
 	err = svc_addsock(nfsd_serv, fd, buf, SIMPLE_TRANSACTION_LIMIT);
-	if (err < 0)
-		lockd_down();
+	if (err < 0) {
+		svc_destroy(nfsd_serv);
+		return err;
+	}
 
-out:
 	/* Decrease the count, but don't shut down the service */
 	nfsd_serv->sv_nrthreads--;
 	return err;
@@ -978,9 +975,6 @@ static ssize_t __write_ports_delfd(char *buf)
 	if (nfsd_serv != NULL)
 		len = svc_sock_names(nfsd_serv, buf,
 					SIMPLE_TRANSACTION_LIMIT, toclose);
-	if (len >= 0)
-		lockd_down();
-
 	kfree(toclose);
 	return len;
 }
@@ -1014,6 +1008,9 @@ static ssize_t __write_ports_addxprt(char *buf)
 				PF_INET6, port, SVC_SOCK_ANONYMOUS);
 	if (err < 0 && err != -EAFNOSUPPORT)
 		goto out_close;
+
+	/* Decrease the count, but don't shut down the service */
+	nfsd_serv->sv_nrthreads--;
 	return 0;
 out_close:
 	xprt = svc_find_xprt(nfsd_serv, transport, PF_INET, port);
@@ -1022,8 +1019,7 @@ out_close:
 		svc_xprt_put(xprt);
 	}
 out_err:
-	/* Decrease the count, but don't shut down the service */
-	nfsd_serv->sv_nrthreads--;
+	svc_destroy(nfsd_serv);
 	return err;
 }
 
@@ -1194,7 +1190,7 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size)
 			bsize = NFSSVC_MAXBLKSIZE;
 		bsize &= ~(1024-1);
 		mutex_lock(&nfsd_mutex);
-		if (nfsd_serv && nfsd_serv->sv_nrthreads) {
+		if (nfsd_serv) {
 			mutex_unlock(&nfsd_mutex);
 			return -EBUSY;
 		}
@@ -1310,6 +1306,8 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size)
 			return -EINVAL;
 
 		status = nfs4_reset_recoverydir(recdir);
+		if (status)
+			return status;
 	}
 
 	return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%s\n",
diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index 72377761270..b76ac3a82e3 100644
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -153,6 +153,7 @@ void		nfsd_lockd_shutdown(void);
 #define nfserr_bad_seqid	cpu_to_be32(NFSERR_BAD_SEQID)
 #define	nfserr_symlink		cpu_to_be32(NFSERR_SYMLINK)
 #define	nfserr_not_same		cpu_to_be32(NFSERR_NOT_SAME)
+#define nfserr_lock_range	cpu_to_be32(NFSERR_LOCK_RANGE)
 #define	nfserr_restorefh	cpu_to_be32(NFSERR_RESTOREFH)
 #define	nfserr_attrnotsupp	cpu_to_be32(NFSERR_ATTRNOTSUPP)
 #define	nfserr_bad_xdr		cpu_to_be32(NFSERR_BAD_XDR)
diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c
index a047ad6111e..08e17264784 100644
--- a/fs/nfsd/nfsproc.c
+++ b/fs/nfsd/nfsproc.c
@@ -144,7 +144,7 @@ nfsd_proc_read(struct svc_rqst *rqstp, struct nfsd_readargs *argp,
 	svc_reserve_auth(rqstp, (19<<2) + argp->count + 4);
 
 	resp->count = argp->count;
-	nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh), NULL,
+	nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh),
 				  argp->offset,
 			   	  rqstp->rq_vec, argp->vlen,
 				  &resp->count);
@@ -290,7 +290,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp,
 	 * gospel of sun micro
 	 */
 	if (type != S_IFREG) {
-		int	is_borc = 0;
 		if (type != S_IFBLK && type != S_IFCHR) {
 			rdev = 0;
 		} else if (type == S_IFCHR && !(attr->ia_valid & ATTR_SIZE)) {
@@ -298,7 +297,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp,
 			type = S_IFIFO;
 		} else {
 			/* Okay, char or block special */
-			is_borc = 1;
 			if (!rdev)
 				rdev = wanted;
 		}
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 06b2a26edfe..e2c43464f23 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -180,15 +180,80 @@ int nfsd_nrthreads(void)
 	return rv;
 }
 
+static int nfsd_init_socks(int port)
+{
+	int error;
+	if (!list_empty(&nfsd_serv->sv_permsocks))
+		return 0;
+
+	error = svc_create_xprt(nfsd_serv, "udp", PF_INET, port,
+					SVC_SOCK_DEFAULTS);
+	if (error < 0)
+		return error;
+
+	error = svc_create_xprt(nfsd_serv, "tcp", PF_INET, port,
+					SVC_SOCK_DEFAULTS);
+	if (error < 0)
+		return error;
+
+	return 0;
+}
+
+static bool nfsd_up = false;
+
+static int nfsd_startup(unsigned short port, int nrservs)
+{
+	int ret;
+
+	if (nfsd_up)
+		return 0;
+	/*
+	 * Readahead param cache - will no-op if it already exists.
+	 * (Note therefore results will be suboptimal if number of
+	 * threads is modified after nfsd start.)
+	 */
+	ret = nfsd_racache_init(2*nrservs);
+	if (ret)
+		return ret;
+	ret = nfsd_init_socks(port);
+	if (ret)
+		goto out_racache;
+	ret = lockd_up();
+	if (ret)
+		goto out_racache;
+	ret = nfs4_state_start();
+	if (ret)
+		goto out_lockd;
+	nfsd_up = true;
+	return 0;
+out_lockd:
+	lockd_down();
+out_racache:
+	nfsd_racache_shutdown();
+	return ret;
+}
+
+static void nfsd_shutdown(void)
+{
+	/*
+	 * write_ports can create the server without actually starting
+	 * any threads--if we get shut down before any threads are
+	 * started, then nfsd_last_thread will be run before any of this
+	 * other initialization has been done.
+	 */
+	if (!nfsd_up)
+		return;
+	nfs4_state_shutdown();
+	lockd_down();
+	nfsd_racache_shutdown();
+	nfsd_up = false;
+}
+
 static void nfsd_last_thread(struct svc_serv *serv)
 {
 	/* When last nfsd thread exits we need to do some clean-up */
-	struct svc_xprt *xprt;
-	list_for_each_entry(xprt, &serv->sv_permsocks, xpt_list)
-		lockd_down();
 	nfsd_serv = NULL;
-	nfsd_racache_shutdown();
-	nfs4_state_shutdown();
+	nfsd_shutdown();
 
 	printk(KERN_WARNING "nfsd: last server has exited, flushing export "
 			    "cache\n");
@@ -263,45 +328,18 @@ int nfsd_create_serv(void)
 		       nfsd_max_blksize >= 8*1024*2)
 			nfsd_max_blksize /= 2;
 	}
+	nfsd_reset_versions();
 
 	nfsd_serv = svc_create_pooled(&nfsd_program, nfsd_max_blksize,
 				      nfsd_last_thread, nfsd, THIS_MODULE);
 	if (nfsd_serv == NULL)
-		err = -ENOMEM;
-	else
-		set_max_drc();
+		return -ENOMEM;
 
+	set_max_drc();
 	do_gettimeofday(&nfssvc_boot);		/* record boot time */
 	return err;
 }
 
-static int nfsd_init_socks(int port)
-{
-	int error;
-	if (!list_empty(&nfsd_serv->sv_permsocks))
-		return 0;
-
-	error = svc_create_xprt(nfsd_serv, "udp", PF_INET, port,
-					SVC_SOCK_DEFAULTS);
-	if (error < 0)
-		return error;
-
-	error = lockd_up();
-	if (error < 0)
-		return error;
-
-	error = svc_create_xprt(nfsd_serv, "tcp", PF_INET, port,
-					SVC_SOCK_DEFAULTS);
-	if (error < 0)
-		return error;
-
-	error = lockd_up();
-	if (error < 0)
-		return error;
-
-	return 0;
-}
-
 int nfsd_nrpools(void)
 {
 	if (nfsd_serv == NULL)
@@ -376,10 +414,16 @@ int nfsd_set_nrthreads(int n, int *nthreads)
 	return err;
 }
 
+/*
+ * Adjust the number of threads and return the new number of threads.
+ * This is also the function that starts the server if necessary, if
+ * this is the first time nrservs is nonzero.
+ */
 int
 nfsd_svc(unsigned short port, int nrservs)
 {
 	int	error;
+	bool	nfsd_up_before;
 
 	mutex_lock(&nfsd_mutex);
 	dprintk("nfsd: creating service\n");
@@ -391,34 +435,29 @@ nfsd_svc(unsigned short port, int nrservs)
 	if (nrservs == 0 && nfsd_serv == NULL)
 		goto out;
 
-	/* Readahead param cache - will no-op if it already exists */
-	error =	nfsd_racache_init(2*nrservs);
-	if (error<0)
-		goto out;
-	error = nfs4_state_start();
+	error = nfsd_create_serv();
 	if (error)
 		goto out;
 
-	nfsd_reset_versions();
-
-	error = nfsd_create_serv();
+	nfsd_up_before = nfsd_up;
 
+	error = nfsd_startup(port, nrservs);
 	if (error)
-		goto out;
-	error = nfsd_init_socks(port);
-	if (error)
-		goto failure;
-
+		goto out_destroy;
 	error = svc_set_num_threads(nfsd_serv, NULL, nrservs);
-	if (error == 0)
-		/* We are holding a reference to nfsd_serv which
-		 * we don't want to count in the return value,
-		 * so subtract 1
-		 */
-		error = nfsd_serv->sv_nrthreads - 1;
- failure:
+	if (error)
+		goto out_shutdown;
+	/* We are holding a reference to nfsd_serv which
+	 * we don't want to count in the return value,
+	 * so subtract 1
+	 */
+	error = nfsd_serv->sv_nrthreads - 1;
+out_shutdown:
+	if (error < 0 && !nfsd_up_before)
+		nfsd_shutdown();
+out_destroy:
 	svc_destroy(nfsd_serv);		/* Release server */
- out:
+out:
 	mutex_unlock(&nfsd_mutex);
 	return error;
 }
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index 006c84230c7..7731a75971d 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -88,7 +88,6 @@ struct nfs4_delegation {
 	struct nfs4_client	*dl_client;
 	struct nfs4_file	*dl_file;
 	struct file_lock	*dl_flock;
-	struct file		*dl_vfs_file;
 	u32			dl_type;
 	time_t			dl_time;
 /* For recall: */
@@ -342,12 +341,50 @@ struct nfs4_file {
 	struct list_head        fi_hash;    /* hash by "struct inode *" */
 	struct list_head        fi_stateids;
 	struct list_head	fi_delegations;
+	/* One each for O_RDONLY, O_WRONLY, O_RDWR: */
+	struct file *		fi_fds[3];
+	/* One each for O_RDONLY, O_WRONLY: */
+	atomic_t		fi_access[2];
+	/*
+	 * Each open stateid contributes 1 to either fi_readers or
+	 * fi_writers, or both, depending on the open mode.  A
+	 * delegation also takes an fi_readers reference.  Lock
+	 * stateid's take none.
+	 */
+	atomic_t		fi_readers;
+	atomic_t		fi_writers;
 	struct inode		*fi_inode;
 	u32                     fi_id;      /* used with stateowner->so_id 
 					     * for stateid_hashtbl hash */
 	bool			fi_had_conflict;
 };
 
+/* XXX: for first cut may fall back on returning file that doesn't work
+ * at all? */
+static inline struct file *find_writeable_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_RDWR])
+		return f->fi_fds[O_RDWR];
+	return f->fi_fds[O_WRONLY];
+}
+
+static inline struct file *find_readable_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_RDWR])
+		return f->fi_fds[O_RDWR];
+	return f->fi_fds[O_RDONLY];
+}
+
+static inline struct file *find_any_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_RDWR])
+		return f->fi_fds[O_RDWR];
+	else if (f->fi_fds[O_RDWR])
+		return f->fi_fds[O_WRONLY];
+	else
+		return f->fi_fds[O_RDONLY];
+}
+
 /*
 * nfs4_stateid can either be an open stateid or (eventually) a lock stateid
 *
@@ -373,7 +410,6 @@ struct nfs4_stateid {
 	struct nfs4_stateowner      * st_stateowner;
 	struct nfs4_file            * st_file;
 	stateid_t                     st_stateid;
-	struct file                 * st_vfs_file;
 	unsigned long                 st_access_bmap;
 	unsigned long                 st_deny_bmap;
 	struct nfs4_stateid         * st_openstp;
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 3c111120b61..9df85a13af2 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -604,7 +604,7 @@ nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry, struct nfs4_ac
 	return error;
 }
 
-#endif /* defined(CONFIG_NFS_V4) */
+#endif /* defined(CONFIG_NFSD_V4) */
 
 #ifdef CONFIG_NFSD_V3
 /*
@@ -903,7 +903,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
               loff_t offset, struct kvec *vec, int vlen, unsigned long *count)
 {
 	struct inode *inode;
-	struct raparms	*ra;
 	mm_segment_t	oldfs;
 	__be32		err;
 	int		host_err;
@@ -914,12 +913,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 	if (svc_msnfs(fhp) && !lock_may_read(inode, offset, *count))
 		goto out;
 
-	/* Get readahead parameters */
-	ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino);
-
-	if (ra && ra->p_set)
-		file->f_ra = ra->p_ra;
-
 	if (file->f_op->splice_read && rqstp->rq_splice_ok) {
 		struct splice_desc sd = {
 			.len		= 0,
@@ -937,16 +930,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		set_fs(oldfs);
 	}
 
-	/* Write back readahead params */
-	if (ra) {
-		struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex];
-		spin_lock(&rab->pb_lock);
-		ra->p_ra = file->f_ra;
-		ra->p_set = 1;
-		ra->p_count--;
-		spin_unlock(&rab->pb_lock);
-	}
-
 	if (host_err >= 0) {
 		nfsdstats.io_read += host_err;
 		*count = host_err;
@@ -1086,8 +1069,45 @@ out:
  * on entry. On return, *count contains the number of bytes actually read.
  * N.B. After this call fhp needs an fh_put
  */
+__be32 nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp,
+	loff_t offset, struct kvec *vec, int vlen, unsigned long *count)
+{
+	struct file *file;
+	struct inode *inode;
+	struct raparms	*ra;
+	__be32 err;
+
+	err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
+	if (err)
+		return err;
+
+	inode = file->f_path.dentry->d_inode;
+
+	/* Get readahead parameters */
+	ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino);
+
+	if (ra && ra->p_set)
+		file->f_ra = ra->p_ra;
+
+	err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
+
+	/* Write back readahead params */
+	if (ra) {
+		struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex];
+		spin_lock(&rab->pb_lock);
+		ra->p_ra = file->f_ra;
+		ra->p_set = 1;
+		ra->p_count--;
+		spin_unlock(&rab->pb_lock);
+	}
+
+	nfsd_close(file);
+	return err;
+}
+
+/* As above, but use the provided file descriptor. */
 __be32
-nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
+nfsd_read_file(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		loff_t offset, struct kvec *vec, int vlen,
 		unsigned long *count)
 {
@@ -1099,13 +1119,8 @@ nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		if (err)
 			goto out;
 		err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
-	} else {
-		err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
-		if (err)
-			goto out;
-		err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
-		nfsd_close(file);
-	}
+	} else /* Note file may still be NULL in NFSv4 special stateid case: */
+		err = nfsd_read(rqstp, fhp, offset, vec, vlen, count);
 out:
 	return err;
 }
@@ -1631,7 +1646,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 				char *name, int len, struct svc_fh *tfhp)
 {
 	struct dentry	*ddir, *dnew, *dold;
-	struct inode	*dirp, *dest;
+	struct inode	*dirp;
 	__be32		err;
 	int		host_err;
 
@@ -1659,7 +1674,6 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 		goto out_nfserr;
 
 	dold = tfhp->fh_dentry;
-	dest = dold->d_inode;
 
 	host_err = mnt_want_write(tfhp->fh_export->ex_path.mnt);
 	if (host_err) {
@@ -2038,7 +2052,6 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
 					struct dentry *dentry, int acc)
 {
 	struct inode	*inode = dentry->d_inode;
-	struct path	path;
 	int		err;
 
 	if (acc == NFSD_MAY_NOP)
@@ -2111,15 +2124,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
 	if (err == -EACCES && S_ISREG(inode->i_mode) &&
 	    acc == (NFSD_MAY_READ | NFSD_MAY_OWNER_OVERRIDE))
 		err = inode_permission(inode, MAY_EXEC);
-	if (err)
-		goto nfsd_out;
 
-	/* Do integrity (permission) checking now, but defer incrementing
-	 * IMA counts to the actual file open.
-	 */
-	path.mnt = exp->ex_path.mnt;
-	path.dentry = dentry;
-nfsd_out:
 	return err? nfserrno(err) : 0;
 }
 
diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h
index 217a62c2a35..9a370a5e36b 100644
--- a/fs/nfsd/vfs.h
+++ b/fs/nfsd/vfs.h
@@ -64,7 +64,9 @@ __be32		nfsd_commit(struct svc_rqst *, struct svc_fh *,
 __be32		nfsd_open(struct svc_rqst *, struct svc_fh *, int,
 				int, struct file **);
 void		nfsd_close(struct file *);
-__be32 		nfsd_read(struct svc_rqst *, struct svc_fh *, struct file *,
+__be32 		nfsd_read(struct svc_rqst *, struct svc_fh *,
+				loff_t, struct kvec *, int, unsigned long *);
+__be32 		nfsd_read_file(struct svc_rqst *, struct svc_fh *, struct file *,
 				loff_t, struct kvec *, int, unsigned long *);
 __be32 		nfsd_write(struct svc_rqst *, struct svc_fh *,struct file *,
 				loff_t, struct kvec *,int, unsigned long *, int *);
diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c
index effdbdbe6c1..3dbdc1d356b 100644
--- a/fs/nilfs2/bmap.c
+++ b/fs/nilfs2/bmap.c
@@ -26,6 +26,8 @@
 #include "nilfs.h"
 #include "bmap.h"
 #include "sb.h"
+#include "btree.h"
+#include "direct.h"
 #include "btnode.h"
 #include "mdt.h"
 #include "dat.h"
@@ -533,7 +535,7 @@ void nilfs_bmap_init_gc(struct nilfs_bmap *bmap)
 
 void nilfs_bmap_init_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap)
 {
-	memcpy(gcbmap, bmap, sizeof(union nilfs_bmap_union));
+	memcpy(gcbmap, bmap, sizeof(*bmap));
 	init_rwsem(&gcbmap->b_sem);
 	lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key);
 	gcbmap->b_inode = &NILFS_BMAP_I(gcbmap)->vfs_inode;
@@ -541,7 +543,7 @@ void nilfs_bmap_init_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap)
 
 void nilfs_bmap_commit_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap)
 {
-	memcpy(bmap, gcbmap, sizeof(union nilfs_bmap_union));
+	memcpy(bmap, gcbmap, sizeof(*bmap));
 	init_rwsem(&bmap->b_sem);
 	lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key);
 	bmap->b_inode = &NILFS_BMAP_I(bmap)->vfs_inode;
diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h
index 9980d7dbab9..a20569b1992 100644
--- a/fs/nilfs2/bmap.h
+++ b/fs/nilfs2/bmap.h
@@ -32,11 +32,6 @@
 
 #define NILFS_BMAP_INVALID_PTR	0
 
-#define nilfs_bmap_dkey_to_key(dkey)	le64_to_cpu(dkey)
-#define nilfs_bmap_key_to_dkey(key)	cpu_to_le64(key)
-#define nilfs_bmap_dptr_to_ptr(dptr)	le64_to_cpu(dptr)
-#define nilfs_bmap_ptr_to_dptr(ptr)	cpu_to_le64(ptr)
-
 #define nilfs_bmap_keydiff_abs(diff)	((diff) < 0 ? -(diff) : (diff))
 
 
@@ -71,7 +66,7 @@ struct nilfs_bmap_operations {
 	int (*bop_delete)(struct nilfs_bmap *, __u64);
 	void (*bop_clear)(struct nilfs_bmap *);
 
-	int (*bop_propagate)(const struct nilfs_bmap *, struct buffer_head *);
+	int (*bop_propagate)(struct nilfs_bmap *, struct buffer_head *);
 	void (*bop_lookup_dirty_buffers)(struct nilfs_bmap *,
 					 struct list_head *);
 
@@ -110,6 +105,7 @@ static inline int nilfs_bmap_is_new_ptr(unsigned long ptr)
  * @b_last_allocated_ptr: last allocated ptr for data block
  * @b_ptr_type: pointer type
  * @b_state: state
+ * @b_nchildren_per_block: maximum number of child nodes for non-root nodes
  */
 struct nilfs_bmap {
 	union {
@@ -123,6 +119,7 @@ struct nilfs_bmap {
 	__u64 b_last_allocated_ptr;
 	int b_ptr_type;
 	int b_state;
+	__u16 b_nchildren_per_block;
 };
 
 /* pointer type */
@@ -224,6 +221,13 @@ static inline void nilfs_bmap_abort_end_ptr(struct nilfs_bmap *bmap,
 		nilfs_dat_abort_end(dat, &req->bpr_req);
 }
 
+static inline void nilfs_bmap_set_target_v(struct nilfs_bmap *bmap, __u64 key,
+					   __u64 ptr)
+{
+	bmap->b_last_allocated_key = key;
+	bmap->b_last_allocated_ptr = ptr;
+}
+
 __u64 nilfs_bmap_data_get_key(const struct nilfs_bmap *,
 			      const struct buffer_head *);
 
diff --git a/fs/nilfs2/bmap_union.h b/fs/nilfs2/bmap_union.h
deleted file mode 100644
index d41509bff47..00000000000
--- a/fs/nilfs2/bmap_union.h
+++ /dev/null
@@ -1,42 +0,0 @@
-/*
- * bmap_union.h - NILFS block mapping.
- *
- * Copyright (C) 2006-2008 Nippon Telegraph and Telephone Corporation.
- *
- * 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.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- *
- * Written by Koji Sato <koji@osrg.net>.
- */
-
-#ifndef _NILFS_BMAP_UNION_H
-#define _NILFS_BMAP_UNION_H
-
-#include "bmap.h"
-#include "direct.h"
-#include "btree.h"
-
-/**
- * nilfs_bmap_union -
- * @bi_bmap: bmap structure
- * @bi_btree: direct map structure
- * @bi_direct: B-tree structure
- */
-union nilfs_bmap_union {
-	struct nilfs_bmap bi_bmap;
-	struct nilfs_direct bi_direct;
-	struct nilfs_btree bi_btree;
-};
-
-#endif	/* _NILFS_BMAP_UNION_H */
diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c
index 447ce47a330..f78ab1044d1 100644
--- a/fs/nilfs2/btnode.c
+++ b/fs/nilfs2/btnode.c
@@ -96,10 +96,12 @@ nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr)
 }
 
 int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
-			      sector_t pblocknr, struct buffer_head **pbh)
+			      sector_t pblocknr, int mode,
+			      struct buffer_head **pbh, sector_t *submit_ptr)
 {
 	struct buffer_head *bh;
 	struct inode *inode = NILFS_BTNC_I(btnc);
+	struct page *page;
 	int err;
 
 	bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node);
@@ -107,6 +109,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
 		return -ENOMEM;
 
 	err = -EEXIST; /* internal code */
+	page = bh->b_page;
 
 	if (buffer_uptodate(bh) || buffer_dirty(bh))
 		goto found;
@@ -125,7 +128,16 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
 			}
 		}
 	}
-	lock_buffer(bh);
+
+	if (mode == READA) {
+		if (pblocknr != *submit_ptr + 1 || !trylock_buffer(bh)) {
+			err = -EBUSY; /* internal code */
+			brelse(bh);
+			goto out_locked;
+		}
+	} else { /* mode == READ */
+		lock_buffer(bh);
+	}
 	if (buffer_uptodate(bh)) {
 		unlock_buffer(bh);
 		err = -EEXIST; /* internal code */
@@ -136,15 +148,16 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
 	bh->b_blocknr = pblocknr; /* set block address for read */
 	bh->b_end_io = end_buffer_read_sync;
 	get_bh(bh);
-	submit_bh(READ, bh);
+	submit_bh(mode, bh);
 	bh->b_blocknr = blocknr; /* set back to the given block address */
+	*submit_ptr = pblocknr;
 	err = 0;
 found:
 	*pbh = bh;
 
 out_locked:
-	unlock_page(bh->b_page);
-	page_cache_release(bh->b_page);
+	unlock_page(page);
+	page_cache_release(page);
 	return err;
 }
 
diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h
index 07da83f0771..79037494f1e 100644
--- a/fs/nilfs2/btnode.h
+++ b/fs/nilfs2/btnode.h
@@ -42,8 +42,8 @@ void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *);
 void nilfs_btnode_cache_clear(struct address_space *);
 struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc,
 					      __u64 blocknr);
-int nilfs_btnode_submit_block(struct address_space *, __u64, sector_t,
-			      struct buffer_head **);
+int nilfs_btnode_submit_block(struct address_space *, __u64, sector_t, int,
+			      struct buffer_head **, sector_t *);
 void nilfs_btnode_delete(struct buffer_head *);
 int nilfs_btnode_prepare_change_key(struct address_space *,
 				    struct nilfs_btnode_chkey_ctxt *);
diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c
index b27a342c5af..300c2bc00c3 100644
--- a/fs/nilfs2/btree.c
+++ b/fs/nilfs2/btree.c
@@ -66,30 +66,10 @@ static void nilfs_btree_free_path(struct nilfs_btree_path *path)
 /*
  * B-tree node operations
  */
-static int nilfs_btree_get_block(const struct nilfs_btree *btree, __u64 ptr,
-				 struct buffer_head **bhp)
-{
-	struct address_space *btnc =
-		&NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache;
-	int err;
-
-	err = nilfs_btnode_submit_block(btnc, ptr, 0, bhp);
-	if (err)
-		return err == -EEXIST ? 0 : err;
-
-	wait_on_buffer(*bhp);
-	if (!buffer_uptodate(*bhp)) {
-		brelse(*bhp);
-		return -EIO;
-	}
-	return 0;
-}
-
-static int nilfs_btree_get_new_block(const struct nilfs_btree *btree,
+static int nilfs_btree_get_new_block(const struct nilfs_bmap *btree,
 				     __u64 ptr, struct buffer_head **bhp)
 {
-	struct address_space *btnc =
-		&NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache;
+	struct address_space *btnc = &NILFS_BMAP_I(btree)->i_btnode_cache;
 	struct buffer_head *bh;
 
 	bh = nilfs_btnode_create_block(btnc, ptr);
@@ -101,71 +81,55 @@ static int nilfs_btree_get_new_block(const struct nilfs_btree *btree,
 	return 0;
 }
 
-static inline int
-nilfs_btree_node_get_flags(const struct nilfs_btree_node *node)
+static int nilfs_btree_node_get_flags(const struct nilfs_btree_node *node)
 {
 	return node->bn_flags;
 }
 
-static inline void
+static void
 nilfs_btree_node_set_flags(struct nilfs_btree_node *node, int flags)
 {
 	node->bn_flags = flags;
 }
 
-static inline int nilfs_btree_node_root(const struct nilfs_btree_node *node)
+static int nilfs_btree_node_root(const struct nilfs_btree_node *node)
 {
 	return nilfs_btree_node_get_flags(node) & NILFS_BTREE_NODE_ROOT;
 }
 
-static inline int
-nilfs_btree_node_get_level(const struct nilfs_btree_node *node)
+static int nilfs_btree_node_get_level(const struct nilfs_btree_node *node)
 {
 	return node->bn_level;
 }
 
-static inline void
+static void
 nilfs_btree_node_set_level(struct nilfs_btree_node *node, int level)
 {
 	node->bn_level = level;
 }
 
-static inline int
-nilfs_btree_node_get_nchildren(const struct nilfs_btree_node *node)
+static int nilfs_btree_node_get_nchildren(const struct nilfs_btree_node *node)
 {
 	return le16_to_cpu(node->bn_nchildren);
 }
 
-static inline void
+static void
 nilfs_btree_node_set_nchildren(struct nilfs_btree_node *node, int nchildren)
 {
 	node->bn_nchildren = cpu_to_le16(nchildren);
 }
 
-static inline int nilfs_btree_node_size(const struct nilfs_btree *btree)
+static int nilfs_btree_node_size(const struct nilfs_bmap *btree)
 {
-	return 1 << btree->bt_bmap.b_inode->i_blkbits;
+	return 1 << btree->b_inode->i_blkbits;
 }
 
-static inline int
-nilfs_btree_node_nchildren_min(const struct nilfs_btree_node *node,
-			       const struct nilfs_btree *btree)
+static int nilfs_btree_nchildren_per_block(const struct nilfs_bmap *btree)
 {
-	return nilfs_btree_node_root(node) ?
-		NILFS_BTREE_ROOT_NCHILDREN_MIN :
-		NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree));
+	return btree->b_nchildren_per_block;
 }
 
-static inline int
-nilfs_btree_node_nchildren_max(const struct nilfs_btree_node *node,
-			       const struct nilfs_btree *btree)
-{
-	return nilfs_btree_node_root(node) ?
-		NILFS_BTREE_ROOT_NCHILDREN_MAX :
-		NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(btree));
-}
-
-static inline __le64 *
+static __le64 *
 nilfs_btree_node_dkeys(const struct nilfs_btree_node *node)
 {
 	return (__le64 *)((char *)(node + 1) +
@@ -173,45 +137,40 @@ nilfs_btree_node_dkeys(const struct nilfs_btree_node *node)
 			   0 : NILFS_BTREE_NODE_EXTRA_PAD_SIZE));
 }
 
-static inline __le64 *
-nilfs_btree_node_dptrs(const struct nilfs_btree_node *node,
-		       const struct nilfs_btree *btree)
+static __le64 *
+nilfs_btree_node_dptrs(const struct nilfs_btree_node *node, int ncmax)
 {
-	return (__le64 *)(nilfs_btree_node_dkeys(node) +
-			  nilfs_btree_node_nchildren_max(node, btree));
+	return (__le64 *)(nilfs_btree_node_dkeys(node) + ncmax);
 }
 
-static inline __u64
+static __u64
 nilfs_btree_node_get_key(const struct nilfs_btree_node *node, int index)
 {
-	return nilfs_bmap_dkey_to_key(*(nilfs_btree_node_dkeys(node) + index));
+	return le64_to_cpu(*(nilfs_btree_node_dkeys(node) + index));
 }
 
-static inline void
+static void
 nilfs_btree_node_set_key(struct nilfs_btree_node *node, int index, __u64 key)
 {
-	*(nilfs_btree_node_dkeys(node) + index) = nilfs_bmap_key_to_dkey(key);
+	*(nilfs_btree_node_dkeys(node) + index) = cpu_to_le64(key);
 }
 
-static inline __u64
-nilfs_btree_node_get_ptr(const struct nilfs_btree *btree,
-			 const struct nilfs_btree_node *node, int index)
+static __u64
+nilfs_btree_node_get_ptr(const struct nilfs_btree_node *node, int index,
+			 int ncmax)
 {
-	return nilfs_bmap_dptr_to_ptr(*(nilfs_btree_node_dptrs(node, btree) +
-					index));
+	return le64_to_cpu(*(nilfs_btree_node_dptrs(node, ncmax) + index));
 }
 
-static inline void
-nilfs_btree_node_set_ptr(struct nilfs_btree *btree,
-			 struct nilfs_btree_node *node, int index, __u64 ptr)
+static void
+nilfs_btree_node_set_ptr(struct nilfs_btree_node *node, int index, __u64 ptr,
+			 int ncmax)
 {
-	*(nilfs_btree_node_dptrs(node, btree) + index) =
-		nilfs_bmap_ptr_to_dptr(ptr);
+	*(nilfs_btree_node_dptrs(node, ncmax) + index) = cpu_to_le64(ptr);
 }
 
-static void nilfs_btree_node_init(struct nilfs_btree *btree,
-				  struct nilfs_btree_node *node,
-				  int flags, int level, int nchildren,
+static void nilfs_btree_node_init(struct nilfs_btree_node *node, int flags,
+				  int level, int nchildren, int ncmax,
 				  const __u64 *keys, const __u64 *ptrs)
 {
 	__le64 *dkeys;
@@ -223,29 +182,28 @@ static void nilfs_btree_node_init(struct nilfs_btree *btree,
 	nilfs_btree_node_set_nchildren(node, nchildren);
 
 	dkeys = nilfs_btree_node_dkeys(node);
-	dptrs = nilfs_btree_node_dptrs(node, btree);
+	dptrs = nilfs_btree_node_dptrs(node, ncmax);
 	for (i = 0; i < nchildren; i++) {
-		dkeys[i] = nilfs_bmap_key_to_dkey(keys[i]);
-		dptrs[i] = nilfs_bmap_ptr_to_dptr(ptrs[i]);
+		dkeys[i] = cpu_to_le64(keys[i]);
+		dptrs[i] = cpu_to_le64(ptrs[i]);
 	}
 }
 
 /* Assume the buffer heads corresponding to left and right are locked. */
-static void nilfs_btree_node_move_left(struct nilfs_btree *btree,
-				       struct nilfs_btree_node *left,
+static void nilfs_btree_node_move_left(struct nilfs_btree_node *left,
 				       struct nilfs_btree_node *right,
-				       int n)
+				       int n, int lncmax, int rncmax)
 {
 	__le64 *ldkeys, *rdkeys;
 	__le64 *ldptrs, *rdptrs;
 	int lnchildren, rnchildren;
 
 	ldkeys = nilfs_btree_node_dkeys(left);
-	ldptrs = nilfs_btree_node_dptrs(left, btree);
+	ldptrs = nilfs_btree_node_dptrs(left, lncmax);
 	lnchildren = nilfs_btree_node_get_nchildren(left);
 
 	rdkeys = nilfs_btree_node_dkeys(right);
-	rdptrs = nilfs_btree_node_dptrs(right, btree);
+	rdptrs = nilfs_btree_node_dptrs(right, rncmax);
 	rnchildren = nilfs_btree_node_get_nchildren(right);
 
 	memcpy(ldkeys + lnchildren, rdkeys, n * sizeof(*rdkeys));
@@ -260,21 +218,20 @@ static void nilfs_btree_node_move_left(struct nilfs_btree *btree,
 }
 
 /* Assume that the buffer heads corresponding to left and right are locked. */
-static void nilfs_btree_node_move_right(struct nilfs_btree *btree,
-					struct nilfs_btree_node *left,
+static void nilfs_btree_node_move_right(struct nilfs_btree_node *left,
 					struct nilfs_btree_node *right,
-					int n)
+					int n, int lncmax, int rncmax)
 {
 	__le64 *ldkeys, *rdkeys;
 	__le64 *ldptrs, *rdptrs;
 	int lnchildren, rnchildren;
 
 	ldkeys = nilfs_btree_node_dkeys(left);
-	ldptrs = nilfs_btree_node_dptrs(left, btree);
+	ldptrs = nilfs_btree_node_dptrs(left, lncmax);
 	lnchildren = nilfs_btree_node_get_nchildren(left);
 
 	rdkeys = nilfs_btree_node_dkeys(right);
-	rdptrs = nilfs_btree_node_dptrs(right, btree);
+	rdptrs = nilfs_btree_node_dptrs(right, rncmax);
 	rnchildren = nilfs_btree_node_get_nchildren(right);
 
 	memmove(rdkeys + n, rdkeys, rnchildren * sizeof(*rdkeys));
@@ -289,16 +246,15 @@ static void nilfs_btree_node_move_right(struct nilfs_btree *btree,
 }
 
 /* Assume that the buffer head corresponding to node is locked. */
-static void nilfs_btree_node_insert(struct nilfs_btree *btree,
-				    struct nilfs_btree_node *node,
-				    __u64 key, __u64 ptr, int index)
+static void nilfs_btree_node_insert(struct nilfs_btree_node *node, int index,
+				    __u64 key, __u64 ptr, int ncmax)
 {
 	__le64 *dkeys;
 	__le64 *dptrs;
 	int nchildren;
 
 	dkeys = nilfs_btree_node_dkeys(node);
-	dptrs = nilfs_btree_node_dptrs(node, btree);
+	dptrs = nilfs_btree_node_dptrs(node, ncmax);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	if (index < nchildren) {
 		memmove(dkeys + index + 1, dkeys + index,
@@ -306,16 +262,15 @@ static void nilfs_btree_node_insert(struct nilfs_btree *btree,
 		memmove(dptrs + index + 1, dptrs + index,
 			(nchildren - index) * sizeof(*dptrs));
 	}
-	dkeys[index] = nilfs_bmap_key_to_dkey(key);
-	dptrs[index] = nilfs_bmap_ptr_to_dptr(ptr);
+	dkeys[index] = cpu_to_le64(key);
+	dptrs[index] = cpu_to_le64(ptr);
 	nchildren++;
 	nilfs_btree_node_set_nchildren(node, nchildren);
 }
 
 /* Assume that the buffer head corresponding to node is locked. */
-static void nilfs_btree_node_delete(struct nilfs_btree *btree,
-				    struct nilfs_btree_node *node,
-				    __u64 *keyp, __u64 *ptrp, int index)
+static void nilfs_btree_node_delete(struct nilfs_btree_node *node, int index,
+				    __u64 *keyp, __u64 *ptrp, int ncmax)
 {
 	__u64 key;
 	__u64 ptr;
@@ -324,9 +279,9 @@ static void nilfs_btree_node_delete(struct nilfs_btree *btree,
 	int nchildren;
 
 	dkeys = nilfs_btree_node_dkeys(node);
-	dptrs = nilfs_btree_node_dptrs(node, btree);
-	key = nilfs_bmap_dkey_to_key(dkeys[index]);
-	ptr = nilfs_bmap_dptr_to_ptr(dptrs[index]);
+	dptrs = nilfs_btree_node_dptrs(node, ncmax);
+	key = le64_to_cpu(dkeys[index]);
+	ptr = le64_to_cpu(dptrs[index]);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	if (keyp != NULL)
 		*keyp = key;
@@ -382,40 +337,92 @@ static int nilfs_btree_node_lookup(const struct nilfs_btree_node *node,
 	return s == 0;
 }
 
-static inline struct nilfs_btree_node *
-nilfs_btree_get_root(const struct nilfs_btree *btree)
+/**
+ * nilfs_btree_node_broken - verify consistency of btree node
+ * @node: btree node block to be examined
+ * @size: node size (in bytes)
+ * @blocknr: block number
+ *
+ * Return Value: If node is broken, 1 is returned. Otherwise, 0 is returned.
+ */
+static int nilfs_btree_node_broken(const struct nilfs_btree_node *node,
+				   size_t size, sector_t blocknr)
 {
-	return (struct nilfs_btree_node *)btree->bt_bmap.b_u.u_data;
+	int level, flags, nchildren;
+	int ret = 0;
+
+	level = nilfs_btree_node_get_level(node);
+	flags = nilfs_btree_node_get_flags(node);
+	nchildren = nilfs_btree_node_get_nchildren(node);
+
+	if (unlikely(level < NILFS_BTREE_LEVEL_NODE_MIN ||
+		     level >= NILFS_BTREE_LEVEL_MAX ||
+		     (flags & NILFS_BTREE_NODE_ROOT) ||
+		     nchildren < 0 ||
+		     nchildren > NILFS_BTREE_NODE_NCHILDREN_MAX(size))) {
+		printk(KERN_CRIT "NILFS: bad btree node (blocknr=%llu): "
+		       "level = %d, flags = 0x%x, nchildren = %d\n",
+		       (unsigned long long)blocknr, level, flags, nchildren);
+		ret = 1;
+	}
+	return ret;
 }
 
-static inline struct nilfs_btree_node *
+int nilfs_btree_broken_node_block(struct buffer_head *bh)
+{
+	int ret;
+
+	if (buffer_nilfs_checked(bh))
+		return 0;
+
+	ret = nilfs_btree_node_broken((struct nilfs_btree_node *)bh->b_data,
+				       bh->b_size, bh->b_blocknr);
+	if (likely(!ret))
+		set_buffer_nilfs_checked(bh);
+	return ret;
+}
+
+static struct nilfs_btree_node *
+nilfs_btree_get_root(const struct nilfs_bmap *btree)
+{
+	return (struct nilfs_btree_node *)btree->b_u.u_data;
+}
+
+static struct nilfs_btree_node *
 nilfs_btree_get_nonroot_node(const struct nilfs_btree_path *path, int level)
 {
 	return (struct nilfs_btree_node *)path[level].bp_bh->b_data;
 }
 
-static inline struct nilfs_btree_node *
+static struct nilfs_btree_node *
 nilfs_btree_get_sib_node(const struct nilfs_btree_path *path, int level)
 {
 	return (struct nilfs_btree_node *)path[level].bp_sib_bh->b_data;
 }
 
-static inline int nilfs_btree_height(const struct nilfs_btree *btree)
+static int nilfs_btree_height(const struct nilfs_bmap *btree)
 {
 	return nilfs_btree_node_get_level(nilfs_btree_get_root(btree)) + 1;
 }
 
-static inline struct nilfs_btree_node *
-nilfs_btree_get_node(const struct nilfs_btree *btree,
+static struct nilfs_btree_node *
+nilfs_btree_get_node(const struct nilfs_bmap *btree,
 		     const struct nilfs_btree_path *path,
-		     int level)
+		     int level, int *ncmaxp)
 {
-	return (level == nilfs_btree_height(btree) - 1) ?
-		nilfs_btree_get_root(btree) :
-		nilfs_btree_get_nonroot_node(path, level);
+	struct nilfs_btree_node *node;
+
+	if (level == nilfs_btree_height(btree) - 1) {
+		node = nilfs_btree_get_root(btree);
+		*ncmaxp = NILFS_BTREE_ROOT_NCHILDREN_MAX;
+	} else {
+		node = nilfs_btree_get_nonroot_node(path, level);
+		*ncmaxp = nilfs_btree_nchildren_per_block(btree);
+	}
+	return node;
 }
 
-static inline int
+static int
 nilfs_btree_bad_node(struct nilfs_btree_node *node, int level)
 {
 	if (unlikely(nilfs_btree_node_get_level(node) != level)) {
@@ -427,13 +434,83 @@ nilfs_btree_bad_node(struct nilfs_btree_node *node, int level)
 	return 0;
 }
 
-static int nilfs_btree_do_lookup(const struct nilfs_btree *btree,
+struct nilfs_btree_readahead_info {
+	struct nilfs_btree_node *node;	/* parent node */
+	int max_ra_blocks;		/* max nof blocks to read ahead */
+	int index;			/* current index on the parent node */
+	int ncmax;			/* nof children in the parent node */
+};
+
+static int __nilfs_btree_get_block(const struct nilfs_bmap *btree, __u64 ptr,
+				   struct buffer_head **bhp,
+				   const struct nilfs_btree_readahead_info *ra)
+{
+	struct address_space *btnc = &NILFS_BMAP_I(btree)->i_btnode_cache;
+	struct buffer_head *bh, *ra_bh;
+	sector_t submit_ptr = 0;
+	int ret;
+
+	ret = nilfs_btnode_submit_block(btnc, ptr, 0, READ, &bh, &submit_ptr);
+	if (ret) {
+		if (ret != -EEXIST)
+			return ret;
+		goto out_check;
+	}
+
+	if (ra) {
+		int i, n;
+		__u64 ptr2;
+
+		/* read ahead sibling nodes */
+		for (n = ra->max_ra_blocks, i = ra->index + 1;
+		     n > 0 && i < ra->ncmax; n--, i++) {
+			ptr2 = nilfs_btree_node_get_ptr(ra->node, i, ra->ncmax);
+
+			ret = nilfs_btnode_submit_block(btnc, ptr2, 0, READA,
+							&ra_bh, &submit_ptr);
+			if (likely(!ret || ret == -EEXIST))
+				brelse(ra_bh);
+			else if (ret != -EBUSY)
+				break;
+			if (!buffer_locked(bh))
+				goto out_no_wait;
+		}
+	}
+
+	wait_on_buffer(bh);
+
+ out_no_wait:
+	if (!buffer_uptodate(bh)) {
+		brelse(bh);
+		return -EIO;
+	}
+
+ out_check:
+	if (nilfs_btree_broken_node_block(bh)) {
+		clear_buffer_uptodate(bh);
+		brelse(bh);
+		return -EINVAL;
+	}
+
+	*bhp = bh;
+	return 0;
+}
+
+static int nilfs_btree_get_block(const struct nilfs_bmap *btree, __u64 ptr,
+				   struct buffer_head **bhp)
+{
+	return __nilfs_btree_get_block(btree, ptr, bhp, NULL);
+}
+
+static int nilfs_btree_do_lookup(const struct nilfs_bmap *btree,
 				 struct nilfs_btree_path *path,
-				 __u64 key, __u64 *ptrp, int minlevel)
+				 __u64 key, __u64 *ptrp, int minlevel,
+				 int readahead)
 {
 	struct nilfs_btree_node *node;
+	struct nilfs_btree_readahead_info p, *ra;
 	__u64 ptr;
-	int level, index, found, ret;
+	int level, index, found, ncmax, ret;
 
 	node = nilfs_btree_get_root(btree);
 	level = nilfs_btree_node_get_level(node);
@@ -441,14 +518,27 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree,
 		return -ENOENT;
 
 	found = nilfs_btree_node_lookup(node, key, &index);
-	ptr = nilfs_btree_node_get_ptr(btree, node, index);
+	ptr = nilfs_btree_node_get_ptr(node, index,
+				       NILFS_BTREE_ROOT_NCHILDREN_MAX);
 	path[level].bp_bh = NULL;
 	path[level].bp_index = index;
 
-	for (level--; level >= minlevel; level--) {
-		ret = nilfs_btree_get_block(btree, ptr, &path[level].bp_bh);
+	ncmax = nilfs_btree_nchildren_per_block(btree);
+
+	while (--level >= minlevel) {
+		ra = NULL;
+		if (level == NILFS_BTREE_LEVEL_NODE_MIN && readahead) {
+			p.node = nilfs_btree_get_node(btree, path, level + 1,
+						      &p.ncmax);
+			p.index = index;
+			p.max_ra_blocks = 7;
+			ra = &p;
+		}
+		ret = __nilfs_btree_get_block(btree, ptr, &path[level].bp_bh,
+					      ra);
 		if (ret < 0)
 			return ret;
+
 		node = nilfs_btree_get_nonroot_node(path, level);
 		if (nilfs_btree_bad_node(node, level))
 			return -EINVAL;
@@ -456,9 +546,9 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree,
 			found = nilfs_btree_node_lookup(node, key, &index);
 		else
 			index = 0;
-		if (index < nilfs_btree_node_nchildren_max(node, btree))
-			ptr = nilfs_btree_node_get_ptr(btree, node, index);
-		else {
+		if (index < ncmax) {
+			ptr = nilfs_btree_node_get_ptr(node, index, ncmax);
+		} else {
 			WARN_ON(found || level != NILFS_BTREE_LEVEL_NODE_MIN);
 			/* insert */
 			ptr = NILFS_BMAP_INVALID_PTR;
@@ -474,22 +564,24 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree,
 	return 0;
 }
 
-static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree,
+static int nilfs_btree_do_lookup_last(const struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
 				      __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node;
 	__u64 ptr;
-	int index, level, ret;
+	int index, level, ncmax, ret;
 
 	node = nilfs_btree_get_root(btree);
 	index = nilfs_btree_node_get_nchildren(node) - 1;
 	if (index < 0)
 		return -ENOENT;
 	level = nilfs_btree_node_get_level(node);
-	ptr = nilfs_btree_node_get_ptr(btree, node, index);
+	ptr = nilfs_btree_node_get_ptr(node, index,
+				       NILFS_BTREE_ROOT_NCHILDREN_MAX);
 	path[level].bp_bh = NULL;
 	path[level].bp_index = index;
+	ncmax = nilfs_btree_nchildren_per_block(btree);
 
 	for (level--; level > 0; level--) {
 		ret = nilfs_btree_get_block(btree, ptr, &path[level].bp_bh);
@@ -499,7 +591,7 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree,
 		if (nilfs_btree_bad_node(node, level))
 			return -EINVAL;
 		index = nilfs_btree_node_get_nchildren(node) - 1;
-		ptr = nilfs_btree_node_get_ptr(btree, node, index);
+		ptr = nilfs_btree_node_get_ptr(node, index, ncmax);
 		path[level].bp_index = index;
 	}
 
@@ -511,51 +603,45 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree,
 	return 0;
 }
 
-static int nilfs_btree_lookup(const struct nilfs_bmap *bmap,
+static int nilfs_btree_lookup(const struct nilfs_bmap *btree,
 			      __u64 key, int level, __u64 *ptrp)
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
-	__u64 ptr;
 	int ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
 
-	ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level);
-
-	if (ptrp != NULL)
-		*ptrp = ptr;
+	ret = nilfs_btree_do_lookup(btree, path, key, ptrp, level, 0);
 
 	nilfs_btree_free_path(path);
 
 	return ret;
 }
 
-static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap,
+static int nilfs_btree_lookup_contig(const struct nilfs_bmap *btree,
 				     __u64 key, __u64 *ptrp, unsigned maxblocks)
 {
-	struct nilfs_btree *btree = (struct nilfs_btree *)bmap;
 	struct nilfs_btree_path *path;
 	struct nilfs_btree_node *node;
 	struct inode *dat = NULL;
 	__u64 ptr, ptr2;
 	sector_t blocknr;
 	int level = NILFS_BTREE_LEVEL_NODE_MIN;
-	int ret, cnt, index, maxlevel;
+	int ret, cnt, index, maxlevel, ncmax;
+	struct nilfs_btree_readahead_info p;
 
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
 
-	ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level);
+	ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level, 1);
 	if (ret < 0)
 		goto out;
 
-	if (NILFS_BMAP_USE_VBN(bmap)) {
-		dat = nilfs_bmap_get_dat(bmap);
+	if (NILFS_BMAP_USE_VBN(btree)) {
+		dat = nilfs_bmap_get_dat(btree);
 		ret = nilfs_dat_translate(dat, ptr, &blocknr);
 		if (ret < 0)
 			goto out;
@@ -566,14 +652,14 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap,
 		goto end;
 
 	maxlevel = nilfs_btree_height(btree) - 1;
-	node = nilfs_btree_get_node(btree, path, level);
+	node = nilfs_btree_get_node(btree, path, level, &ncmax);
 	index = path[level].bp_index + 1;
 	for (;;) {
 		while (index < nilfs_btree_node_get_nchildren(node)) {
 			if (nilfs_btree_node_get_key(node, index) !=
 			    key + cnt)
 				goto end;
-			ptr2 = nilfs_btree_node_get_ptr(btree, node, index);
+			ptr2 = nilfs_btree_node_get_ptr(node, index, ncmax);
 			if (dat) {
 				ret = nilfs_dat_translate(dat, ptr2, &blocknr);
 				if (ret < 0)
@@ -589,20 +675,24 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap,
 			break;
 
 		/* look-up right sibling node */
-		node = nilfs_btree_get_node(btree, path, level + 1);
-		index = path[level + 1].bp_index + 1;
-		if (index >= nilfs_btree_node_get_nchildren(node) ||
-		    nilfs_btree_node_get_key(node, index) != key + cnt)
+		p.node = nilfs_btree_get_node(btree, path, level + 1, &p.ncmax);
+		p.index = path[level + 1].bp_index + 1;
+		p.max_ra_blocks = 7;
+		if (p.index >= nilfs_btree_node_get_nchildren(p.node) ||
+		    nilfs_btree_node_get_key(p.node, p.index) != key + cnt)
 			break;
-		ptr2 = nilfs_btree_node_get_ptr(btree, node, index);
-		path[level + 1].bp_index = index;
+		ptr2 = nilfs_btree_node_get_ptr(p.node, p.index, p.ncmax);
+		path[level + 1].bp_index = p.index;
 
 		brelse(path[level].bp_bh);
 		path[level].bp_bh = NULL;
-		ret = nilfs_btree_get_block(btree, ptr2, &path[level].bp_bh);
+
+		ret = __nilfs_btree_get_block(btree, ptr2, &path[level].bp_bh,
+					      &p);
 		if (ret < 0)
 			goto out;
 		node = nilfs_btree_get_nonroot_node(path, level);
+		ncmax = nilfs_btree_nchildren_per_block(btree);
 		index = 0;
 		path[level].bp_index = index;
 	}
@@ -614,7 +704,7 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap,
 	return ret;
 }
 
-static void nilfs_btree_promote_key(struct nilfs_btree *btree,
+static void nilfs_btree_promote_key(struct nilfs_bmap *btree,
 				    struct nilfs_btree_path *path,
 				    int level, __u64 key)
 {
@@ -636,16 +726,18 @@ static void nilfs_btree_promote_key(struct nilfs_btree *btree,
 	}
 }
 
-static void nilfs_btree_do_insert(struct nilfs_btree *btree,
+static void nilfs_btree_do_insert(struct nilfs_bmap *btree,
 				  struct nilfs_btree_path *path,
 				  int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node;
+	int ncblk;
 
 	if (level < nilfs_btree_height(btree) - 1) {
 		node = nilfs_btree_get_nonroot_node(path, level);
-		nilfs_btree_node_insert(btree, node, *keyp, *ptrp,
-					path[level].bp_index);
+		ncblk = nilfs_btree_nchildren_per_block(btree);
+		nilfs_btree_node_insert(node, path[level].bp_index,
+					*keyp, *ptrp, ncblk);
 		if (!buffer_dirty(path[level].bp_bh))
 			nilfs_btnode_mark_dirty(path[level].bp_bh);
 
@@ -655,22 +747,24 @@ static void nilfs_btree_do_insert(struct nilfs_btree *btree,
 									 0));
 	} else {
 		node = nilfs_btree_get_root(btree);
-		nilfs_btree_node_insert(btree, node, *keyp, *ptrp,
-					path[level].bp_index);
+		nilfs_btree_node_insert(node, path[level].bp_index,
+					*keyp, *ptrp,
+					NILFS_BTREE_ROOT_NCHILDREN_MAX);
 	}
 }
 
-static void nilfs_btree_carry_left(struct nilfs_btree *btree,
+static void nilfs_btree_carry_left(struct nilfs_bmap *btree,
 				   struct nilfs_btree_path *path,
 				   int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *left;
-	int nchildren, lnchildren, n, move;
+	int nchildren, lnchildren, n, move, ncblk;
 
 	node = nilfs_btree_get_nonroot_node(path, level);
 	left = nilfs_btree_get_sib_node(path, level);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	lnchildren = nilfs_btree_node_get_nchildren(left);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 	move = 0;
 
 	n = (nchildren + lnchildren + 1) / 2 - lnchildren;
@@ -680,7 +774,7 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree,
 		move = 1;
 	}
 
-	nilfs_btree_node_move_left(btree, left, node, n);
+	nilfs_btree_node_move_left(left, node, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -705,17 +799,18 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree,
 	nilfs_btree_do_insert(btree, path, level, keyp, ptrp);
 }
 
-static void nilfs_btree_carry_right(struct nilfs_btree *btree,
+static void nilfs_btree_carry_right(struct nilfs_bmap *btree,
 				    struct nilfs_btree_path *path,
 				    int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *right;
-	int nchildren, rnchildren, n, move;
+	int nchildren, rnchildren, n, move, ncblk;
 
 	node = nilfs_btree_get_nonroot_node(path, level);
 	right = nilfs_btree_get_sib_node(path, level);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	rnchildren = nilfs_btree_node_get_nchildren(right);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 	move = 0;
 
 	n = (nchildren + rnchildren + 1) / 2 - rnchildren;
@@ -725,7 +820,7 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree,
 		move = 1;
 	}
 
-	nilfs_btree_node_move_right(btree, node, right, n);
+	nilfs_btree_node_move_right(node, right, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -751,18 +846,19 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree,
 	nilfs_btree_do_insert(btree, path, level, keyp, ptrp);
 }
 
-static void nilfs_btree_split(struct nilfs_btree *btree,
+static void nilfs_btree_split(struct nilfs_bmap *btree,
 			      struct nilfs_btree_path *path,
 			      int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *right;
 	__u64 newkey;
 	__u64 newptr;
-	int nchildren, n, move;
+	int nchildren, n, move, ncblk;
 
 	node = nilfs_btree_get_nonroot_node(path, level);
 	right = nilfs_btree_get_sib_node(path, level);
 	nchildren = nilfs_btree_node_get_nchildren(node);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 	move = 0;
 
 	n = (nchildren + 1) / 2;
@@ -771,7 +867,7 @@ static void nilfs_btree_split(struct nilfs_btree *btree,
 		move = 1;
 	}
 
-	nilfs_btree_node_move_right(btree, node, right, n);
+	nilfs_btree_node_move_right(node, right, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -783,8 +879,8 @@ static void nilfs_btree_split(struct nilfs_btree *btree,
 
 	if (move) {
 		path[level].bp_index -= nilfs_btree_node_get_nchildren(node);
-		nilfs_btree_node_insert(btree, right, *keyp, *ptrp,
-					path[level].bp_index);
+		nilfs_btree_node_insert(right, path[level].bp_index,
+					*keyp, *ptrp, ncblk);
 
 		*keyp = nilfs_btree_node_get_key(right, 0);
 		*ptrp = path[level].bp_newreq.bpr_ptr;
@@ -805,19 +901,21 @@ static void nilfs_btree_split(struct nilfs_btree *btree,
 	path[level + 1].bp_index++;
 }
 
-static void nilfs_btree_grow(struct nilfs_btree *btree,
+static void nilfs_btree_grow(struct nilfs_bmap *btree,
 			     struct nilfs_btree_path *path,
 			     int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *root, *child;
-	int n;
+	int n, ncblk;
 
 	root = nilfs_btree_get_root(btree);
 	child = nilfs_btree_get_sib_node(path, level);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
 	n = nilfs_btree_node_get_nchildren(root);
 
-	nilfs_btree_node_move_right(btree, root, child, n);
+	nilfs_btree_node_move_right(root, child, n,
+				    NILFS_BTREE_ROOT_NCHILDREN_MAX, ncblk);
 	nilfs_btree_node_set_level(root, level + 1);
 
 	if (!buffer_dirty(path[level].bp_sib_bh))
@@ -832,11 +930,11 @@ static void nilfs_btree_grow(struct nilfs_btree *btree,
 	*ptrp = path[level].bp_newreq.bpr_ptr;
 }
 
-static __u64 nilfs_btree_find_near(const struct nilfs_btree *btree,
+static __u64 nilfs_btree_find_near(const struct nilfs_bmap *btree,
 				   const struct nilfs_btree_path *path)
 {
 	struct nilfs_btree_node *node;
-	int level;
+	int level, ncmax;
 
 	if (path == NULL)
 		return NILFS_BMAP_INVALID_PTR;
@@ -844,29 +942,30 @@ static __u64 nilfs_btree_find_near(const struct nilfs_btree *btree,
 	/* left sibling */
 	level = NILFS_BTREE_LEVEL_NODE_MIN;
 	if (path[level].bp_index > 0) {
-		node = nilfs_btree_get_node(btree, path, level);
-		return nilfs_btree_node_get_ptr(btree, node,
-						path[level].bp_index - 1);
+		node = nilfs_btree_get_node(btree, path, level, &ncmax);
+		return nilfs_btree_node_get_ptr(node,
+						path[level].bp_index - 1,
+						ncmax);
 	}
 
 	/* parent */
 	level = NILFS_BTREE_LEVEL_NODE_MIN + 1;
 	if (level <= nilfs_btree_height(btree) - 1) {
-		node = nilfs_btree_get_node(btree, path, level);
-		return nilfs_btree_node_get_ptr(btree, node,
-						path[level].bp_index);
+		node = nilfs_btree_get_node(btree, path, level, &ncmax);
+		return nilfs_btree_node_get_ptr(node, path[level].bp_index,
+						ncmax);
 	}
 
 	return NILFS_BMAP_INVALID_PTR;
 }
 
-static __u64 nilfs_btree_find_target_v(const struct nilfs_btree *btree,
+static __u64 nilfs_btree_find_target_v(const struct nilfs_bmap *btree,
 				       const struct nilfs_btree_path *path,
 				       __u64 key)
 {
 	__u64 ptr;
 
-	ptr = nilfs_bmap_find_target_seq(&btree->bt_bmap, key);
+	ptr = nilfs_bmap_find_target_seq(btree, key);
 	if (ptr != NILFS_BMAP_INVALID_PTR)
 		/* sequential access */
 		return ptr;
@@ -877,17 +976,10 @@ static __u64 nilfs_btree_find_target_v(const struct nilfs_btree *btree,
 			return ptr;
 	}
 	/* block group */
-	return nilfs_bmap_find_target_in_group(&btree->bt_bmap);
-}
-
-static void nilfs_btree_set_target_v(struct nilfs_btree *btree, __u64 key,
-				     __u64 ptr)
-{
-	btree->bt_bmap.b_last_allocated_key = key;
-	btree->bt_bmap.b_last_allocated_ptr = ptr;
+	return nilfs_bmap_find_target_in_group(btree);
 }
 
-static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
+static int nilfs_btree_prepare_insert(struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
 				      int *levelp, __u64 key, __u64 ptr,
 				      struct nilfs_bmap_stats *stats)
@@ -895,79 +987,78 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 	struct buffer_head *bh;
 	struct nilfs_btree_node *node, *parent, *sib;
 	__u64 sibptr;
-	int pindex, level, ret;
+	int pindex, level, ncmax, ncblk, ret;
 	struct inode *dat = NULL;
 
 	stats->bs_nblocks = 0;
 	level = NILFS_BTREE_LEVEL_DATA;
 
 	/* allocate a new ptr for data block */
-	if (NILFS_BMAP_USE_VBN(&btree->bt_bmap)) {
+	if (NILFS_BMAP_USE_VBN(btree)) {
 		path[level].bp_newreq.bpr_ptr =
 			nilfs_btree_find_target_v(btree, path, key);
-		dat = nilfs_bmap_get_dat(&btree->bt_bmap);
+		dat = nilfs_bmap_get_dat(btree);
 	}
 
-	ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap,
-					   &path[level].bp_newreq, dat);
+	ret = nilfs_bmap_prepare_alloc_ptr(btree, &path[level].bp_newreq, dat);
 	if (ret < 0)
 		goto err_out_data;
 
+	ncblk = nilfs_btree_nchildren_per_block(btree);
+
 	for (level = NILFS_BTREE_LEVEL_NODE_MIN;
 	     level < nilfs_btree_height(btree) - 1;
 	     level++) {
 		node = nilfs_btree_get_nonroot_node(path, level);
-		if (nilfs_btree_node_get_nchildren(node) <
-		    nilfs_btree_node_nchildren_max(node, btree)) {
+		if (nilfs_btree_node_get_nchildren(node) < ncblk) {
 			path[level].bp_op = nilfs_btree_do_insert;
 			stats->bs_nblocks++;
 			goto out;
 		}
 
-		parent = nilfs_btree_get_node(btree, path, level + 1);
+		parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
 		pindex = path[level + 1].bp_index;
 
 		/* left sibling */
 		if (pindex > 0) {
-			sibptr = nilfs_btree_node_get_ptr(btree, parent,
-							  pindex - 1);
+			sibptr = nilfs_btree_node_get_ptr(parent, pindex - 1,
+							  ncmax);
 			ret = nilfs_btree_get_block(btree, sibptr, &bh);
 			if (ret < 0)
 				goto err_out_child_node;
 			sib = (struct nilfs_btree_node *)bh->b_data;
-			if (nilfs_btree_node_get_nchildren(sib) <
-			    nilfs_btree_node_nchildren_max(sib, btree)) {
+			if (nilfs_btree_node_get_nchildren(sib) < ncblk) {
 				path[level].bp_sib_bh = bh;
 				path[level].bp_op = nilfs_btree_carry_left;
 				stats->bs_nblocks++;
 				goto out;
-			} else
+			} else {
 				brelse(bh);
+			}
 		}
 
 		/* right sibling */
-		if (pindex <
-		    nilfs_btree_node_get_nchildren(parent) - 1) {
-			sibptr = nilfs_btree_node_get_ptr(btree, parent,
-							  pindex + 1);
+		if (pindex < nilfs_btree_node_get_nchildren(parent) - 1) {
+			sibptr = nilfs_btree_node_get_ptr(parent, pindex + 1,
+							  ncmax);
 			ret = nilfs_btree_get_block(btree, sibptr, &bh);
 			if (ret < 0)
 				goto err_out_child_node;
 			sib = (struct nilfs_btree_node *)bh->b_data;
-			if (nilfs_btree_node_get_nchildren(sib) <
-			    nilfs_btree_node_nchildren_max(sib, btree)) {
+			if (nilfs_btree_node_get_nchildren(sib) < ncblk) {
 				path[level].bp_sib_bh = bh;
 				path[level].bp_op = nilfs_btree_carry_right;
 				stats->bs_nblocks++;
 				goto out;
-			} else
+			} else {
 				brelse(bh);
+			}
 		}
 
 		/* split */
 		path[level].bp_newreq.bpr_ptr =
 			path[level - 1].bp_newreq.bpr_ptr + 1;
-		ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap,
+		ret = nilfs_bmap_prepare_alloc_ptr(btree,
 						   &path[level].bp_newreq, dat);
 		if (ret < 0)
 			goto err_out_child_node;
@@ -979,9 +1070,8 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 
 		stats->bs_nblocks++;
 
-		nilfs_btree_node_init(btree,
-				      (struct nilfs_btree_node *)bh->b_data,
-				      0, level, 0, NULL, NULL);
+		sib = (struct nilfs_btree_node *)bh->b_data;
+		nilfs_btree_node_init(sib, 0, level, 0, ncblk, NULL, NULL);
 		path[level].bp_sib_bh = bh;
 		path[level].bp_op = nilfs_btree_split;
 	}
@@ -989,7 +1079,7 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 	/* root */
 	node = nilfs_btree_get_root(btree);
 	if (nilfs_btree_node_get_nchildren(node) <
-	    nilfs_btree_node_nchildren_max(node, btree)) {
+	    NILFS_BTREE_ROOT_NCHILDREN_MAX) {
 		path[level].bp_op = nilfs_btree_do_insert;
 		stats->bs_nblocks++;
 		goto out;
@@ -997,8 +1087,7 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 
 	/* grow */
 	path[level].bp_newreq.bpr_ptr = path[level - 1].bp_newreq.bpr_ptr + 1;
-	ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap,
-					   &path[level].bp_newreq, dat);
+	ret = nilfs_bmap_prepare_alloc_ptr(btree, &path[level].bp_newreq, dat);
 	if (ret < 0)
 		goto err_out_child_node;
 	ret = nilfs_btree_get_new_block(btree, path[level].bp_newreq.bpr_ptr,
@@ -1006,8 +1095,8 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 	if (ret < 0)
 		goto err_out_curr_node;
 
-	nilfs_btree_node_init(btree, (struct nilfs_btree_node *)bh->b_data,
-			      0, level, 0, NULL, NULL);
+	nilfs_btree_node_init((struct nilfs_btree_node *)bh->b_data,
+			      0, level, 0, ncblk, NULL, NULL);
 	path[level].bp_sib_bh = bh;
 	path[level].bp_op = nilfs_btree_grow;
 
@@ -1024,25 +1113,22 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree,
 
 	/* error */
  err_out_curr_node:
-	nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap, &path[level].bp_newreq,
-				   dat);
+	nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat);
  err_out_child_node:
 	for (level--; level > NILFS_BTREE_LEVEL_DATA; level--) {
 		nilfs_btnode_delete(path[level].bp_sib_bh);
-		nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap,
-					   &path[level].bp_newreq, dat);
+		nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat);
 
 	}
 
-	nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap, &path[level].bp_newreq,
-				   dat);
+	nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat);
  err_out_data:
 	*levelp = level;
 	stats->bs_nblocks = 0;
 	return ret;
 }
 
-static void nilfs_btree_commit_insert(struct nilfs_btree *btree,
+static void nilfs_btree_commit_insert(struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
 				      int maxlevel, __u64 key, __u64 ptr)
 {
@@ -1051,35 +1137,33 @@ static void nilfs_btree_commit_insert(struct nilfs_btree *btree,
 
 	set_buffer_nilfs_volatile((struct buffer_head *)((unsigned long)ptr));
 	ptr = path[NILFS_BTREE_LEVEL_DATA].bp_newreq.bpr_ptr;
-	if (NILFS_BMAP_USE_VBN(&btree->bt_bmap)) {
-		nilfs_btree_set_target_v(btree, key, ptr);
-		dat = nilfs_bmap_get_dat(&btree->bt_bmap);
+	if (NILFS_BMAP_USE_VBN(btree)) {
+		nilfs_bmap_set_target_v(btree, key, ptr);
+		dat = nilfs_bmap_get_dat(btree);
 	}
 
 	for (level = NILFS_BTREE_LEVEL_NODE_MIN; level <= maxlevel; level++) {
-		nilfs_bmap_commit_alloc_ptr(&btree->bt_bmap,
+		nilfs_bmap_commit_alloc_ptr(btree,
 					    &path[level - 1].bp_newreq, dat);
 		path[level].bp_op(btree, path, level, &key, &ptr);
 	}
 
-	if (!nilfs_bmap_dirty(&btree->bt_bmap))
-		nilfs_bmap_set_dirty(&btree->bt_bmap);
+	if (!nilfs_bmap_dirty(btree))
+		nilfs_bmap_set_dirty(btree);
 }
 
-static int nilfs_btree_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
+static int nilfs_btree_insert(struct nilfs_bmap *btree, __u64 key, __u64 ptr)
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	struct nilfs_bmap_stats stats;
 	int level, ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
 
 	ret = nilfs_btree_do_lookup(btree, path, key, NULL,
-				    NILFS_BTREE_LEVEL_NODE_MIN);
+				    NILFS_BTREE_LEVEL_NODE_MIN, 0);
 	if (ret != -ENOENT) {
 		if (ret == 0)
 			ret = -EEXIST;
@@ -1090,23 +1174,25 @@ static int nilfs_btree_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
 	if (ret < 0)
 		goto out;
 	nilfs_btree_commit_insert(btree, path, level, key, ptr);
-	nilfs_bmap_add_blocks(bmap, stats.bs_nblocks);
+	nilfs_bmap_add_blocks(btree, stats.bs_nblocks);
 
  out:
 	nilfs_btree_free_path(path);
 	return ret;
 }
 
-static void nilfs_btree_do_delete(struct nilfs_btree *btree,
+static void nilfs_btree_do_delete(struct nilfs_bmap *btree,
 				  struct nilfs_btree_path *path,
 				  int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node;
+	int ncblk;
 
 	if (level < nilfs_btree_height(btree) - 1) {
 		node = nilfs_btree_get_nonroot_node(path, level);
-		nilfs_btree_node_delete(btree, node, keyp, ptrp,
-					path[level].bp_index);
+		ncblk = nilfs_btree_nchildren_per_block(btree);
+		nilfs_btree_node_delete(node, path[level].bp_index,
+					keyp, ptrp, ncblk);
 		if (!buffer_dirty(path[level].bp_bh))
 			nilfs_btnode_mark_dirty(path[level].bp_bh);
 		if (path[level].bp_index == 0)
@@ -1114,17 +1200,18 @@ static void nilfs_btree_do_delete(struct nilfs_btree *btree,
 				nilfs_btree_node_get_key(node, 0));
 	} else {
 		node = nilfs_btree_get_root(btree);
-		nilfs_btree_node_delete(btree, node, keyp, ptrp,
-					path[level].bp_index);
+		nilfs_btree_node_delete(node, path[level].bp_index,
+					keyp, ptrp,
+					NILFS_BTREE_ROOT_NCHILDREN_MAX);
 	}
 }
 
-static void nilfs_btree_borrow_left(struct nilfs_btree *btree,
+static void nilfs_btree_borrow_left(struct nilfs_bmap *btree,
 				    struct nilfs_btree_path *path,
 				    int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *left;
-	int nchildren, lnchildren, n;
+	int nchildren, lnchildren, n, ncblk;
 
 	nilfs_btree_do_delete(btree, path, level, keyp, ptrp);
 
@@ -1132,10 +1219,11 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree,
 	left = nilfs_btree_get_sib_node(path, level);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	lnchildren = nilfs_btree_node_get_nchildren(left);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
 	n = (nchildren + lnchildren) / 2 - nchildren;
 
-	nilfs_btree_node_move_right(btree, left, node, n);
+	nilfs_btree_node_move_right(left, node, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -1150,12 +1238,12 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree,
 	path[level].bp_index += n;
 }
 
-static void nilfs_btree_borrow_right(struct nilfs_btree *btree,
+static void nilfs_btree_borrow_right(struct nilfs_bmap *btree,
 				     struct nilfs_btree_path *path,
 				     int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *right;
-	int nchildren, rnchildren, n;
+	int nchildren, rnchildren, n, ncblk;
 
 	nilfs_btree_do_delete(btree, path, level, keyp, ptrp);
 
@@ -1163,10 +1251,11 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree,
 	right = nilfs_btree_get_sib_node(path, level);
 	nchildren = nilfs_btree_node_get_nchildren(node);
 	rnchildren = nilfs_btree_node_get_nchildren(right);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
 	n = (nchildren + rnchildren) / 2 - nchildren;
 
-	nilfs_btree_node_move_left(btree, node, right, n);
+	nilfs_btree_node_move_left(node, right, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -1182,21 +1271,22 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree,
 	path[level].bp_sib_bh = NULL;
 }
 
-static void nilfs_btree_concat_left(struct nilfs_btree *btree,
+static void nilfs_btree_concat_left(struct nilfs_bmap *btree,
 				    struct nilfs_btree_path *path,
 				    int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *left;
-	int n;
+	int n, ncblk;
 
 	nilfs_btree_do_delete(btree, path, level, keyp, ptrp);
 
 	node = nilfs_btree_get_nonroot_node(path, level);
 	left = nilfs_btree_get_sib_node(path, level);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
 	n = nilfs_btree_node_get_nchildren(node);
 
-	nilfs_btree_node_move_left(btree, left, node, n);
+	nilfs_btree_node_move_left(left, node, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_sib_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_sib_bh);
@@ -1207,21 +1297,22 @@ static void nilfs_btree_concat_left(struct nilfs_btree *btree,
 	path[level].bp_index += nilfs_btree_node_get_nchildren(left);
 }
 
-static void nilfs_btree_concat_right(struct nilfs_btree *btree,
+static void nilfs_btree_concat_right(struct nilfs_bmap *btree,
 				     struct nilfs_btree_path *path,
 				     int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *node, *right;
-	int n;
+	int n, ncblk;
 
 	nilfs_btree_do_delete(btree, path, level, keyp, ptrp);
 
 	node = nilfs_btree_get_nonroot_node(path, level);
 	right = nilfs_btree_get_sib_node(path, level);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
 	n = nilfs_btree_node_get_nchildren(right);
 
-	nilfs_btree_node_move_left(btree, node, right, n);
+	nilfs_btree_node_move_left(node, right, n, ncblk, ncblk);
 
 	if (!buffer_dirty(path[level].bp_bh))
 		nilfs_btnode_mark_dirty(path[level].bp_bh);
@@ -1231,29 +1322,32 @@ static void nilfs_btree_concat_right(struct nilfs_btree *btree,
 	path[level + 1].bp_index++;
 }
 
-static void nilfs_btree_shrink(struct nilfs_btree *btree,
+static void nilfs_btree_shrink(struct nilfs_bmap *btree,
 			       struct nilfs_btree_path *path,
 			       int level, __u64 *keyp, __u64 *ptrp)
 {
 	struct nilfs_btree_node *root, *child;
-	int n;
+	int n, ncblk;
 
 	nilfs_btree_do_delete(btree, path, level, keyp, ptrp);
 
 	root = nilfs_btree_get_root(btree);
 	child = nilfs_btree_get_nonroot_node(path, level);
+	ncblk = nilfs_btree_nchildren_per_block(btree);
 
-	nilfs_btree_node_delete(btree, root, NULL, NULL, 0);
+	nilfs_btree_node_delete(root, 0, NULL, NULL,
+				NILFS_BTREE_ROOT_NCHILDREN_MAX);
 	nilfs_btree_node_set_level(root, level);
 	n = nilfs_btree_node_get_nchildren(child);
-	nilfs_btree_node_move_left(btree, root, child, n);
+	nilfs_btree_node_move_left(root, child, n,
+				   NILFS_BTREE_ROOT_NCHILDREN_MAX, ncblk);
 
 	nilfs_btnode_delete(path[level].bp_bh);
 	path[level].bp_bh = NULL;
 }
 
 
-static int nilfs_btree_prepare_delete(struct nilfs_btree *btree,
+static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
 				      int *levelp,
 				      struct nilfs_bmap_stats *stats,
@@ -1262,42 +1356,43 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree,
 	struct buffer_head *bh;
 	struct nilfs_btree_node *node, *parent, *sib;
 	__u64 sibptr;
-	int pindex, level, ret;
+	int pindex, level, ncmin, ncmax, ncblk, ret;
 
 	ret = 0;
 	stats->bs_nblocks = 0;
+	ncmin = NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree));
+	ncblk = nilfs_btree_nchildren_per_block(btree);
+
 	for (level = NILFS_BTREE_LEVEL_NODE_MIN;
 	     level < nilfs_btree_height(btree) - 1;
 	     level++) {
 		node = nilfs_btree_get_nonroot_node(path, level);
 		path[level].bp_oldreq.bpr_ptr =
-			nilfs_btree_node_get_ptr(btree, node,
-						 path[level].bp_index);
-		ret = nilfs_bmap_prepare_end_ptr(&btree->bt_bmap,
+			nilfs_btree_node_get_ptr(node, path[level].bp_index,
+						 ncblk);
+		ret = nilfs_bmap_prepare_end_ptr(btree,
 						 &path[level].bp_oldreq, dat);
 		if (ret < 0)
 			goto err_out_child_node;
 
-		if (nilfs_btree_node_get_nchildren(node) >
-		    nilfs_btree_node_nchildren_min(node, btree)) {
+		if (nilfs_btree_node_get_nchildren(node) > ncmin) {
 			path[level].bp_op = nilfs_btree_do_delete;
 			stats->bs_nblocks++;
 			goto out;
 		}
 
-		parent = nilfs_btree_get_node(btree, path, level + 1);
+		parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
 		pindex = path[level + 1].bp_index;
 
 		if (pindex > 0) {
 			/* left sibling */
-			sibptr = nilfs_btree_node_get_ptr(btree, parent,
-							  pindex - 1);
+			sibptr = nilfs_btree_node_get_ptr(parent, pindex - 1,
+							  ncmax);
 			ret = nilfs_btree_get_block(btree, sibptr, &bh);
 			if (ret < 0)
 				goto err_out_curr_node;
 			sib = (struct nilfs_btree_node *)bh->b_data;
-			if (nilfs_btree_node_get_nchildren(sib) >
-			    nilfs_btree_node_nchildren_min(sib, btree)) {
+			if (nilfs_btree_node_get_nchildren(sib) > ncmin) {
 				path[level].bp_sib_bh = bh;
 				path[level].bp_op = nilfs_btree_borrow_left;
 				stats->bs_nblocks++;
@@ -1311,14 +1406,13 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree,
 		} else if (pindex <
 			   nilfs_btree_node_get_nchildren(parent) - 1) {
 			/* right sibling */
-			sibptr = nilfs_btree_node_get_ptr(btree, parent,
-							  pindex + 1);
+			sibptr = nilfs_btree_node_get_ptr(parent, pindex + 1,
+							  ncmax);
 			ret = nilfs_btree_get_block(btree, sibptr, &bh);
 			if (ret < 0)
 				goto err_out_curr_node;
 			sib = (struct nilfs_btree_node *)bh->b_data;
-			if (nilfs_btree_node_get_nchildren(sib) >
-			    nilfs_btree_node_nchildren_min(sib, btree)) {
+			if (nilfs_btree_node_get_nchildren(sib) > ncmin) {
 				path[level].bp_sib_bh = bh;
 				path[level].bp_op = nilfs_btree_borrow_right;
 				stats->bs_nblocks++;
@@ -1349,10 +1443,10 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree,
 
 	node = nilfs_btree_get_root(btree);
 	path[level].bp_oldreq.bpr_ptr =
-		nilfs_btree_node_get_ptr(btree, node, path[level].bp_index);
+		nilfs_btree_node_get_ptr(node, path[level].bp_index,
+					 NILFS_BTREE_ROOT_NCHILDREN_MAX);
 
-	ret = nilfs_bmap_prepare_end_ptr(&btree->bt_bmap,
-					 &path[level].bp_oldreq, dat);
+	ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat);
 	if (ret < 0)
 		goto err_out_child_node;
 
@@ -1367,75 +1461,68 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree,
 
 	/* error */
  err_out_curr_node:
-	nilfs_bmap_abort_end_ptr(&btree->bt_bmap, &path[level].bp_oldreq, dat);
+	nilfs_bmap_abort_end_ptr(btree, &path[level].bp_oldreq, dat);
  err_out_child_node:
 	for (level--; level >= NILFS_BTREE_LEVEL_NODE_MIN; level--) {
 		brelse(path[level].bp_sib_bh);
-		nilfs_bmap_abort_end_ptr(&btree->bt_bmap,
-					 &path[level].bp_oldreq, dat);
+		nilfs_bmap_abort_end_ptr(btree, &path[level].bp_oldreq, dat);
 	}
 	*levelp = level;
 	stats->bs_nblocks = 0;
 	return ret;
 }
 
-static void nilfs_btree_commit_delete(struct nilfs_btree *btree,
+static void nilfs_btree_commit_delete(struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
 				      int maxlevel, struct inode *dat)
 {
 	int level;
 
 	for (level = NILFS_BTREE_LEVEL_NODE_MIN; level <= maxlevel; level++) {
-		nilfs_bmap_commit_end_ptr(&btree->bt_bmap,
-					  &path[level].bp_oldreq, dat);
+		nilfs_bmap_commit_end_ptr(btree, &path[level].bp_oldreq, dat);
 		path[level].bp_op(btree, path, level, NULL, NULL);
 	}
 
-	if (!nilfs_bmap_dirty(&btree->bt_bmap))
-		nilfs_bmap_set_dirty(&btree->bt_bmap);
+	if (!nilfs_bmap_dirty(btree))
+		nilfs_bmap_set_dirty(btree);
 }
 
-static int nilfs_btree_delete(struct nilfs_bmap *bmap, __u64 key)
+static int nilfs_btree_delete(struct nilfs_bmap *btree, __u64 key)
 
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	struct nilfs_bmap_stats stats;
 	struct inode *dat;
 	int level, ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
 
 	ret = nilfs_btree_do_lookup(btree, path, key, NULL,
-				    NILFS_BTREE_LEVEL_NODE_MIN);
+				    NILFS_BTREE_LEVEL_NODE_MIN, 0);
 	if (ret < 0)
 		goto out;
 
 
-	dat = NILFS_BMAP_USE_VBN(&btree->bt_bmap) ?
-		nilfs_bmap_get_dat(&btree->bt_bmap) : NULL;
+	dat = NILFS_BMAP_USE_VBN(btree) ? nilfs_bmap_get_dat(btree) : NULL;
 
 	ret = nilfs_btree_prepare_delete(btree, path, &level, &stats, dat);
 	if (ret < 0)
 		goto out;
 	nilfs_btree_commit_delete(btree, path, level, dat);
-	nilfs_bmap_sub_blocks(bmap, stats.bs_nblocks);
+	nilfs_bmap_sub_blocks(btree, stats.bs_nblocks);
 
 out:
 	nilfs_btree_free_path(path);
 	return ret;
 }
 
-static int nilfs_btree_last_key(const struct nilfs_bmap *bmap, __u64 *keyp)
+static int nilfs_btree_last_key(const struct nilfs_bmap *btree, __u64 *keyp)
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	int ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
@@ -1447,16 +1534,14 @@ static int nilfs_btree_last_key(const struct nilfs_bmap *bmap, __u64 *keyp)
 	return ret;
 }
 
-static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key)
+static int nilfs_btree_check_delete(struct nilfs_bmap *btree, __u64 key)
 {
 	struct buffer_head *bh;
-	struct nilfs_btree *btree;
 	struct nilfs_btree_node *root, *node;
 	__u64 maxkey, nextmaxkey;
 	__u64 ptr;
 	int nchildren, ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	root = nilfs_btree_get_root(btree);
 	switch (nilfs_btree_height(btree)) {
 	case 2:
@@ -1467,7 +1552,8 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key)
 		nchildren = nilfs_btree_node_get_nchildren(root);
 		if (nchildren > 1)
 			return 0;
-		ptr = nilfs_btree_node_get_ptr(btree, root, nchildren - 1);
+		ptr = nilfs_btree_node_get_ptr(root, nchildren - 1,
+					       NILFS_BTREE_ROOT_NCHILDREN_MAX);
 		ret = nilfs_btree_get_block(btree, ptr, &bh);
 		if (ret < 0)
 			return ret;
@@ -1487,32 +1573,33 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key)
 	return (maxkey == key) && (nextmaxkey < NILFS_BMAP_LARGE_LOW);
 }
 
-static int nilfs_btree_gather_data(struct nilfs_bmap *bmap,
+static int nilfs_btree_gather_data(struct nilfs_bmap *btree,
 				   __u64 *keys, __u64 *ptrs, int nitems)
 {
 	struct buffer_head *bh;
-	struct nilfs_btree *btree;
 	struct nilfs_btree_node *node, *root;
 	__le64 *dkeys;
 	__le64 *dptrs;
 	__u64 ptr;
-	int nchildren, i, ret;
+	int nchildren, ncmax, i, ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	root = nilfs_btree_get_root(btree);
 	switch (nilfs_btree_height(btree)) {
 	case 2:
 		bh = NULL;
 		node = root;
+		ncmax = NILFS_BTREE_ROOT_NCHILDREN_MAX;
 		break;
 	case 3:
 		nchildren = nilfs_btree_node_get_nchildren(root);
 		WARN_ON(nchildren > 1);
-		ptr = nilfs_btree_node_get_ptr(btree, root, nchildren - 1);
+		ptr = nilfs_btree_node_get_ptr(root, nchildren - 1,
+					       NILFS_BTREE_ROOT_NCHILDREN_MAX);
 		ret = nilfs_btree_get_block(btree, ptr, &bh);
 		if (ret < 0)
 			return ret;
 		node = (struct nilfs_btree_node *)bh->b_data;
+		ncmax = nilfs_btree_nchildren_per_block(btree);
 		break;
 	default:
 		node = NULL;
@@ -1523,10 +1610,10 @@ static int nilfs_btree_gather_data(struct nilfs_bmap *bmap,
 	if (nchildren < nitems)
 		nitems = nchildren;
 	dkeys = nilfs_btree_node_dkeys(node);
-	dptrs = nilfs_btree_node_dptrs(node, btree);
+	dptrs = nilfs_btree_node_dptrs(node, ncmax);
 	for (i = 0; i < nitems; i++) {
-		keys[i] = nilfs_bmap_dkey_to_key(dkeys[i]);
-		ptrs[i] = nilfs_bmap_dptr_to_ptr(dptrs[i]);
+		keys[i] = le64_to_cpu(dkeys[i]);
+		ptrs[i] = le64_to_cpu(dptrs[i]);
 	}
 
 	if (bh != NULL)
@@ -1536,14 +1623,13 @@ static int nilfs_btree_gather_data(struct nilfs_bmap *bmap,
 }
 
 static int
-nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key,
+nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *btree, __u64 key,
 				       union nilfs_bmap_ptr_req *dreq,
 				       union nilfs_bmap_ptr_req *nreq,
 				       struct buffer_head **bhp,
 				       struct nilfs_bmap_stats *stats)
 {
 	struct buffer_head *bh;
-	struct nilfs_btree *btree = (struct nilfs_btree *)bmap;
 	struct inode *dat = NULL;
 	int ret;
 
@@ -1551,12 +1637,12 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key,
 
 	/* for data */
 	/* cannot find near ptr */
-	if (NILFS_BMAP_USE_VBN(bmap)) {
+	if (NILFS_BMAP_USE_VBN(btree)) {
 		dreq->bpr_ptr = nilfs_btree_find_target_v(btree, NULL, key);
-		dat = nilfs_bmap_get_dat(bmap);
+		dat = nilfs_bmap_get_dat(btree);
 	}
 
-	ret = nilfs_bmap_prepare_alloc_ptr(bmap, dreq, dat);
+	ret = nilfs_bmap_prepare_alloc_ptr(btree, dreq, dat);
 	if (ret < 0)
 		return ret;
 
@@ -1564,7 +1650,7 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key,
 	stats->bs_nblocks++;
 	if (nreq != NULL) {
 		nreq->bpr_ptr = dreq->bpr_ptr + 1;
-		ret = nilfs_bmap_prepare_alloc_ptr(bmap, nreq, dat);
+		ret = nilfs_bmap_prepare_alloc_ptr(btree, nreq, dat);
 		if (ret < 0)
 			goto err_out_dreq;
 
@@ -1581,16 +1667,16 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key,
 
 	/* error */
  err_out_nreq:
-	nilfs_bmap_abort_alloc_ptr(bmap, nreq, dat);
+	nilfs_bmap_abort_alloc_ptr(btree, nreq, dat);
  err_out_dreq:
-	nilfs_bmap_abort_alloc_ptr(bmap, dreq, dat);
+	nilfs_bmap_abort_alloc_ptr(btree, dreq, dat);
 	stats->bs_nblocks = 0;
 	return ret;
 
 }
 
 static void
-nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap,
+nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *btree,
 				      __u64 key, __u64 ptr,
 				      const __u64 *keys, const __u64 *ptrs,
 				      int n,
@@ -1598,57 +1684,59 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap,
 				      union nilfs_bmap_ptr_req *nreq,
 				      struct buffer_head *bh)
 {
-	struct nilfs_btree *btree = (struct nilfs_btree *)bmap;
 	struct nilfs_btree_node *node;
 	struct inode *dat;
 	__u64 tmpptr;
+	int ncblk;
 
 	/* free resources */
-	if (bmap->b_ops->bop_clear != NULL)
-		bmap->b_ops->bop_clear(bmap);
+	if (btree->b_ops->bop_clear != NULL)
+		btree->b_ops->bop_clear(btree);
 
 	/* ptr must be a pointer to a buffer head. */
 	set_buffer_nilfs_volatile((struct buffer_head *)((unsigned long)ptr));
 
 	/* convert and insert */
-	dat = NILFS_BMAP_USE_VBN(bmap) ? nilfs_bmap_get_dat(bmap) : NULL;
-	nilfs_btree_init(bmap);
+	dat = NILFS_BMAP_USE_VBN(btree) ? nilfs_bmap_get_dat(btree) : NULL;
+	nilfs_btree_init(btree);
 	if (nreq != NULL) {
-		nilfs_bmap_commit_alloc_ptr(bmap, dreq, dat);
-		nilfs_bmap_commit_alloc_ptr(bmap, nreq, dat);
+		nilfs_bmap_commit_alloc_ptr(btree, dreq, dat);
+		nilfs_bmap_commit_alloc_ptr(btree, nreq, dat);
 
 		/* create child node at level 1 */
 		node = (struct nilfs_btree_node *)bh->b_data;
-		nilfs_btree_node_init(btree, node, 0, 1, n, keys, ptrs);
-		nilfs_btree_node_insert(btree, node,
-					key, dreq->bpr_ptr, n);
+		ncblk = nilfs_btree_nchildren_per_block(btree);
+		nilfs_btree_node_init(node, 0, 1, n, ncblk, keys, ptrs);
+		nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr, ncblk);
 		if (!buffer_dirty(bh))
 			nilfs_btnode_mark_dirty(bh);
-		if (!nilfs_bmap_dirty(bmap))
-			nilfs_bmap_set_dirty(bmap);
+		if (!nilfs_bmap_dirty(btree))
+			nilfs_bmap_set_dirty(btree);
 
 		brelse(bh);
 
 		/* create root node at level 2 */
 		node = nilfs_btree_get_root(btree);
 		tmpptr = nreq->bpr_ptr;
-		nilfs_btree_node_init(btree, node, NILFS_BTREE_NODE_ROOT,
-				      2, 1, &keys[0], &tmpptr);
+		nilfs_btree_node_init(node, NILFS_BTREE_NODE_ROOT, 2, 1,
+				      NILFS_BTREE_ROOT_NCHILDREN_MAX,
+				      &keys[0], &tmpptr);
 	} else {
-		nilfs_bmap_commit_alloc_ptr(bmap, dreq, dat);
+		nilfs_bmap_commit_alloc_ptr(btree, dreq, dat);
 
 		/* create root node at level 1 */
 		node = nilfs_btree_get_root(btree);
-		nilfs_btree_node_init(btree, node, NILFS_BTREE_NODE_ROOT,
-				      1, n, keys, ptrs);
-		nilfs_btree_node_insert(btree, node,
-					key, dreq->bpr_ptr, n);
-		if (!nilfs_bmap_dirty(bmap))
-			nilfs_bmap_set_dirty(bmap);
+		nilfs_btree_node_init(node, NILFS_BTREE_NODE_ROOT, 1, n,
+				      NILFS_BTREE_ROOT_NCHILDREN_MAX,
+				      keys, ptrs);
+		nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr,
+					NILFS_BTREE_ROOT_NCHILDREN_MAX);
+		if (!nilfs_bmap_dirty(btree))
+			nilfs_bmap_set_dirty(btree);
 	}
 
-	if (NILFS_BMAP_USE_VBN(bmap))
-		nilfs_btree_set_target_v(btree, key, dreq->bpr_ptr);
+	if (NILFS_BMAP_USE_VBN(btree))
+		nilfs_bmap_set_target_v(btree, key, dreq->bpr_ptr);
 }
 
 /**
@@ -1660,7 +1748,7 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap,
  * @ptrs:
  * @n:
  */
-int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap,
+int nilfs_btree_convert_and_insert(struct nilfs_bmap *btree,
 				   __u64 key, __u64 ptr,
 				   const __u64 *keys, const __u64 *ptrs, int n)
 {
@@ -1673,7 +1761,7 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap,
 		di = &dreq;
 		ni = NULL;
 	} else if ((n + 1) <= NILFS_BTREE_NODE_NCHILDREN_MAX(
-			   1 << bmap->b_inode->i_blkbits)) {
+			   1 << btree->b_inode->i_blkbits)) {
 		di = &dreq;
 		ni = &nreq;
 	} else {
@@ -1682,17 +1770,17 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap,
 		BUG();
 	}
 
-	ret = nilfs_btree_prepare_convert_and_insert(bmap, key, di, ni, &bh,
+	ret = nilfs_btree_prepare_convert_and_insert(btree, key, di, ni, &bh,
 						     &stats);
 	if (ret < 0)
 		return ret;
-	nilfs_btree_commit_convert_and_insert(bmap, key, ptr, keys, ptrs, n,
+	nilfs_btree_commit_convert_and_insert(btree, key, ptr, keys, ptrs, n,
 					      di, ni, bh);
-	nilfs_bmap_add_blocks(bmap, stats.bs_nblocks);
+	nilfs_bmap_add_blocks(btree, stats.bs_nblocks);
 	return 0;
 }
 
-static int nilfs_btree_propagate_p(struct nilfs_btree *btree,
+static int nilfs_btree_propagate_p(struct nilfs_bmap *btree,
 				   struct nilfs_btree_path *path,
 				   int level,
 				   struct buffer_head *bh)
@@ -1704,17 +1792,17 @@ static int nilfs_btree_propagate_p(struct nilfs_btree *btree,
 	return 0;
 }
 
-static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree,
+static int nilfs_btree_prepare_update_v(struct nilfs_bmap *btree,
 					struct nilfs_btree_path *path,
 					int level, struct inode *dat)
 {
 	struct nilfs_btree_node *parent;
-	int ret;
+	int ncmax, ret;
 
-	parent = nilfs_btree_get_node(btree, path, level + 1);
+	parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
 	path[level].bp_oldreq.bpr_ptr =
-		nilfs_btree_node_get_ptr(btree, parent,
-					 path[level + 1].bp_index);
+		nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index,
+					 ncmax);
 	path[level].bp_newreq.bpr_ptr = path[level].bp_oldreq.bpr_ptr + 1;
 	ret = nilfs_dat_prepare_update(dat, &path[level].bp_oldreq.bpr_req,
 				       &path[level].bp_newreq.bpr_req);
@@ -1726,7 +1814,7 @@ static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree,
 		path[level].bp_ctxt.newkey = path[level].bp_newreq.bpr_ptr;
 		path[level].bp_ctxt.bh = path[level].bp_bh;
 		ret = nilfs_btnode_prepare_change_key(
-			&NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache,
+			&NILFS_BMAP_I(btree)->i_btnode_cache,
 			&path[level].bp_ctxt);
 		if (ret < 0) {
 			nilfs_dat_abort_update(dat,
@@ -1739,30 +1827,31 @@ static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree,
 	return 0;
 }
 
-static void nilfs_btree_commit_update_v(struct nilfs_btree *btree,
+static void nilfs_btree_commit_update_v(struct nilfs_bmap *btree,
 					struct nilfs_btree_path *path,
 					int level, struct inode *dat)
 {
 	struct nilfs_btree_node *parent;
+	int ncmax;
 
 	nilfs_dat_commit_update(dat, &path[level].bp_oldreq.bpr_req,
 				&path[level].bp_newreq.bpr_req,
-				btree->bt_bmap.b_ptr_type == NILFS_BMAP_PTR_VS);
+				btree->b_ptr_type == NILFS_BMAP_PTR_VS);
 
 	if (buffer_nilfs_node(path[level].bp_bh)) {
 		nilfs_btnode_commit_change_key(
-			&NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache,
+			&NILFS_BMAP_I(btree)->i_btnode_cache,
 			&path[level].bp_ctxt);
 		path[level].bp_bh = path[level].bp_ctxt.bh;
 	}
 	set_buffer_nilfs_volatile(path[level].bp_bh);
 
-	parent = nilfs_btree_get_node(btree, path, level + 1);
-	nilfs_btree_node_set_ptr(btree, parent, path[level + 1].bp_index,
-				 path[level].bp_newreq.bpr_ptr);
+	parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
+	nilfs_btree_node_set_ptr(parent, path[level + 1].bp_index,
+				 path[level].bp_newreq.bpr_ptr, ncmax);
 }
 
-static void nilfs_btree_abort_update_v(struct nilfs_btree *btree,
+static void nilfs_btree_abort_update_v(struct nilfs_bmap *btree,
 				       struct nilfs_btree_path *path,
 				       int level, struct inode *dat)
 {
@@ -1770,11 +1859,11 @@ static void nilfs_btree_abort_update_v(struct nilfs_btree *btree,
 			       &path[level].bp_newreq.bpr_req);
 	if (buffer_nilfs_node(path[level].bp_bh))
 		nilfs_btnode_abort_change_key(
-			&NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache,
+			&NILFS_BMAP_I(btree)->i_btnode_cache,
 			&path[level].bp_ctxt);
 }
 
-static int nilfs_btree_prepare_propagate_v(struct nilfs_btree *btree,
+static int nilfs_btree_prepare_propagate_v(struct nilfs_bmap *btree,
 					   struct nilfs_btree_path *path,
 					   int minlevel, int *maxlevelp,
 					   struct inode *dat)
@@ -1809,7 +1898,7 @@ static int nilfs_btree_prepare_propagate_v(struct nilfs_btree *btree,
 	return ret;
 }
 
-static void nilfs_btree_commit_propagate_v(struct nilfs_btree *btree,
+static void nilfs_btree_commit_propagate_v(struct nilfs_bmap *btree,
 					   struct nilfs_btree_path *path,
 					   int minlevel, int maxlevel,
 					   struct buffer_head *bh,
@@ -1824,14 +1913,15 @@ static void nilfs_btree_commit_propagate_v(struct nilfs_btree *btree,
 		nilfs_btree_commit_update_v(btree, path, level, dat);
 }
 
-static int nilfs_btree_propagate_v(struct nilfs_btree *btree,
+static int nilfs_btree_propagate_v(struct nilfs_bmap *btree,
 				   struct nilfs_btree_path *path,
 				   int level, struct buffer_head *bh)
 {
 	int maxlevel = 0, ret;
 	struct nilfs_btree_node *parent;
-	struct inode *dat = nilfs_bmap_get_dat(&btree->bt_bmap);
+	struct inode *dat = nilfs_bmap_get_dat(btree);
 	__u64 ptr;
+	int ncmax;
 
 	get_bh(bh);
 	path[level].bp_bh = bh;
@@ -1841,9 +1931,10 @@ static int nilfs_btree_propagate_v(struct nilfs_btree *btree,
 		goto out;
 
 	if (buffer_nilfs_volatile(path[level].bp_bh)) {
-		parent = nilfs_btree_get_node(btree, path, level + 1);
-		ptr = nilfs_btree_node_get_ptr(btree, parent,
-					       path[level + 1].bp_index);
+		parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
+		ptr = nilfs_btree_node_get_ptr(parent,
+					       path[level + 1].bp_index,
+					       ncmax);
 		ret = nilfs_dat_mark_dirty(dat, ptr);
 		if (ret < 0)
 			goto out;
@@ -1857,10 +1948,9 @@ static int nilfs_btree_propagate_v(struct nilfs_btree *btree,
 	return ret;
 }
 
-static int nilfs_btree_propagate(const struct nilfs_bmap *bmap,
+static int nilfs_btree_propagate(struct nilfs_bmap *btree,
 				 struct buffer_head *bh)
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	struct nilfs_btree_node *node;
 	__u64 key;
@@ -1868,7 +1958,6 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap,
 
 	WARN_ON(!buffer_dirty(bh));
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
@@ -1878,11 +1967,11 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap,
 		key = nilfs_btree_node_get_key(node, 0);
 		level = nilfs_btree_node_get_level(node);
 	} else {
-		key = nilfs_bmap_data_get_key(bmap, bh);
+		key = nilfs_bmap_data_get_key(btree, bh);
 		level = NILFS_BTREE_LEVEL_DATA;
 	}
 
-	ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1);
+	ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1, 0);
 	if (ret < 0) {
 		if (unlikely(ret == -ENOENT))
 			printk(KERN_CRIT "%s: key = %llu, level == %d\n",
@@ -1890,7 +1979,7 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap,
 		goto out;
 	}
 
-	ret = NILFS_BMAP_USE_VBN(bmap) ?
+	ret = NILFS_BMAP_USE_VBN(btree) ?
 		nilfs_btree_propagate_v(btree, path, level, bh) :
 		nilfs_btree_propagate_p(btree, path, level, bh);
 
@@ -1900,13 +1989,13 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap,
 	return ret;
 }
 
-static int nilfs_btree_propagate_gc(const struct nilfs_bmap *bmap,
+static int nilfs_btree_propagate_gc(struct nilfs_bmap *btree,
 				    struct buffer_head *bh)
 {
-	return nilfs_dat_mark_dirty(nilfs_bmap_get_dat(bmap), bh->b_blocknr);
+	return nilfs_dat_mark_dirty(nilfs_bmap_get_dat(btree), bh->b_blocknr);
 }
 
-static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree,
+static void nilfs_btree_add_dirty_buffer(struct nilfs_bmap *btree,
 					 struct list_head *lists,
 					 struct buffer_head *bh)
 {
@@ -1920,6 +2009,18 @@ static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree,
 	node = (struct nilfs_btree_node *)bh->b_data;
 	key = nilfs_btree_node_get_key(node, 0);
 	level = nilfs_btree_node_get_level(node);
+	if (level < NILFS_BTREE_LEVEL_NODE_MIN ||
+	    level >= NILFS_BTREE_LEVEL_MAX) {
+		dump_stack();
+		printk(KERN_WARNING
+		       "%s: invalid btree level: %d (key=%llu, ino=%lu, "
+		       "blocknr=%llu)\n",
+		       __func__, level, (unsigned long long)key,
+		       NILFS_BMAP_I(btree)->vfs_inode.i_ino,
+		       (unsigned long long)bh->b_blocknr);
+		return;
+	}
+
 	list_for_each(head, &lists[level]) {
 		cbh = list_entry(head, struct buffer_head, b_assoc_buffers);
 		cnode = (struct nilfs_btree_node *)cbh->b_data;
@@ -1930,11 +2031,10 @@ static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree,
 	list_add_tail(&bh->b_assoc_buffers, head);
 }
 
-static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *bmap,
+static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *btree,
 					     struct list_head *listp)
 {
-	struct nilfs_btree *btree = (struct nilfs_btree *)bmap;
-	struct address_space *btcache = &NILFS_BMAP_I(bmap)->i_btnode_cache;
+	struct address_space *btcache = &NILFS_BMAP_I(btree)->i_btnode_cache;
 	struct list_head lists[NILFS_BTREE_LEVEL_MAX];
 	struct pagevec pvec;
 	struct buffer_head *bh, *head;
@@ -1968,7 +2068,7 @@ static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *bmap,
 		list_splice_tail(&lists[level], listp);
 }
 
-static int nilfs_btree_assign_p(struct nilfs_btree *btree,
+static int nilfs_btree_assign_p(struct nilfs_bmap *btree,
 				struct nilfs_btree_path *path,
 				int level,
 				struct buffer_head **bh,
@@ -1978,38 +2078,38 @@ static int nilfs_btree_assign_p(struct nilfs_btree *btree,
 	struct nilfs_btree_node *parent;
 	__u64 key;
 	__u64 ptr;
-	int ret;
+	int ncmax, ret;
 
-	parent = nilfs_btree_get_node(btree, path, level + 1);
-	ptr = nilfs_btree_node_get_ptr(btree, parent,
-				       path[level + 1].bp_index);
+	parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
+	ptr = nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index,
+				       ncmax);
 	if (buffer_nilfs_node(*bh)) {
 		path[level].bp_ctxt.oldkey = ptr;
 		path[level].bp_ctxt.newkey = blocknr;
 		path[level].bp_ctxt.bh = *bh;
 		ret = nilfs_btnode_prepare_change_key(
-			&NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache,
+			&NILFS_BMAP_I(btree)->i_btnode_cache,
 			&path[level].bp_ctxt);
 		if (ret < 0)
 			return ret;
 		nilfs_btnode_commit_change_key(
-			&NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache,
+			&NILFS_BMAP_I(btree)->i_btnode_cache,
 			&path[level].bp_ctxt);
 		*bh = path[level].bp_ctxt.bh;
 	}
 
-	nilfs_btree_node_set_ptr(btree, parent,
-				 path[level + 1].bp_index, blocknr);
+	nilfs_btree_node_set_ptr(parent, path[level + 1].bp_index, blocknr,
+				 ncmax);
 
 	key = nilfs_btree_node_get_key(parent, path[level + 1].bp_index);
 	/* on-disk format */
-	binfo->bi_dat.bi_blkoff = nilfs_bmap_key_to_dkey(key);
+	binfo->bi_dat.bi_blkoff = cpu_to_le64(key);
 	binfo->bi_dat.bi_level = level;
 
 	return 0;
 }
 
-static int nilfs_btree_assign_v(struct nilfs_btree *btree,
+static int nilfs_btree_assign_v(struct nilfs_bmap *btree,
 				struct nilfs_btree_path *path,
 				int level,
 				struct buffer_head **bh,
@@ -2017,15 +2117,15 @@ static int nilfs_btree_assign_v(struct nilfs_btree *btree,
 				union nilfs_binfo *binfo)
 {
 	struct nilfs_btree_node *parent;
-	struct inode *dat = nilfs_bmap_get_dat(&btree->bt_bmap);
+	struct inode *dat = nilfs_bmap_get_dat(btree);
 	__u64 key;
 	__u64 ptr;
 	union nilfs_bmap_ptr_req req;
-	int ret;
+	int ncmax, ret;
 
-	parent = nilfs_btree_get_node(btree, path, level + 1);
-	ptr = nilfs_btree_node_get_ptr(btree, parent,
-				       path[level + 1].bp_index);
+	parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
+	ptr = nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index,
+				       ncmax);
 	req.bpr_ptr = ptr;
 	ret = nilfs_dat_prepare_start(dat, &req.bpr_req);
 	if (ret < 0)
@@ -2034,24 +2134,22 @@ static int nilfs_btree_assign_v(struct nilfs_btree *btree,
 
 	key = nilfs_btree_node_get_key(parent, path[level + 1].bp_index);
 	/* on-disk format */
-	binfo->bi_v.bi_vblocknr = nilfs_bmap_ptr_to_dptr(ptr);
-	binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key);
+	binfo->bi_v.bi_vblocknr = cpu_to_le64(ptr);
+	binfo->bi_v.bi_blkoff = cpu_to_le64(key);
 
 	return 0;
 }
 
-static int nilfs_btree_assign(struct nilfs_bmap *bmap,
+static int nilfs_btree_assign(struct nilfs_bmap *btree,
 			      struct buffer_head **bh,
 			      sector_t blocknr,
 			      union nilfs_binfo *binfo)
 {
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	struct nilfs_btree_node *node;
 	__u64 key;
 	int level, ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
@@ -2061,17 +2159,17 @@ static int nilfs_btree_assign(struct nilfs_bmap *bmap,
 		key = nilfs_btree_node_get_key(node, 0);
 		level = nilfs_btree_node_get_level(node);
 	} else {
-		key = nilfs_bmap_data_get_key(bmap, *bh);
+		key = nilfs_bmap_data_get_key(btree, *bh);
 		level = NILFS_BTREE_LEVEL_DATA;
 	}
 
-	ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1);
+	ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1, 0);
 	if (ret < 0) {
 		WARN_ON(ret == -ENOENT);
 		goto out;
 	}
 
-	ret = NILFS_BMAP_USE_VBN(bmap) ?
+	ret = NILFS_BMAP_USE_VBN(btree) ?
 		nilfs_btree_assign_v(btree, path, level, bh, blocknr, binfo) :
 		nilfs_btree_assign_p(btree, path, level, bh, blocknr, binfo);
 
@@ -2081,7 +2179,7 @@ static int nilfs_btree_assign(struct nilfs_bmap *bmap,
 	return ret;
 }
 
-static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap,
+static int nilfs_btree_assign_gc(struct nilfs_bmap *btree,
 				 struct buffer_head **bh,
 				 sector_t blocknr,
 				 union nilfs_binfo *binfo)
@@ -2090,7 +2188,7 @@ static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap,
 	__u64 key;
 	int ret;
 
-	ret = nilfs_dat_move(nilfs_bmap_get_dat(bmap), (*bh)->b_blocknr,
+	ret = nilfs_dat_move(nilfs_bmap_get_dat(btree), (*bh)->b_blocknr,
 			     blocknr);
 	if (ret < 0)
 		return ret;
@@ -2099,29 +2197,27 @@ static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap,
 		node = (struct nilfs_btree_node *)(*bh)->b_data;
 		key = nilfs_btree_node_get_key(node, 0);
 	} else
-		key = nilfs_bmap_data_get_key(bmap, *bh);
+		key = nilfs_bmap_data_get_key(btree, *bh);
 
 	/* on-disk format */
 	binfo->bi_v.bi_vblocknr = cpu_to_le64((*bh)->b_blocknr);
-	binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key);
+	binfo->bi_v.bi_blkoff = cpu_to_le64(key);
 
 	return 0;
 }
 
-static int nilfs_btree_mark(struct nilfs_bmap *bmap, __u64 key, int level)
+static int nilfs_btree_mark(struct nilfs_bmap *btree, __u64 key, int level)
 {
 	struct buffer_head *bh;
-	struct nilfs_btree *btree;
 	struct nilfs_btree_path *path;
 	__u64 ptr;
 	int ret;
 
-	btree = (struct nilfs_btree *)bmap;
 	path = nilfs_btree_alloc_path();
 	if (path == NULL)
 		return -ENOMEM;
 
-	ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level + 1);
+	ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level + 1, 0);
 	if (ret < 0) {
 		WARN_ON(ret == -ENOENT);
 		goto out;
@@ -2135,8 +2231,8 @@ static int nilfs_btree_mark(struct nilfs_bmap *bmap, __u64 key, int level)
 	if (!buffer_dirty(bh))
 		nilfs_btnode_mark_dirty(bh);
 	brelse(bh);
-	if (!nilfs_bmap_dirty(&btree->bt_bmap))
-		nilfs_bmap_set_dirty(&btree->bt_bmap);
+	if (!nilfs_bmap_dirty(btree))
+		nilfs_bmap_set_dirty(btree);
 
  out:
 	nilfs_btree_free_path(path);
@@ -2186,10 +2282,14 @@ static const struct nilfs_bmap_operations nilfs_btree_ops_gc = {
 int nilfs_btree_init(struct nilfs_bmap *bmap)
 {
 	bmap->b_ops = &nilfs_btree_ops;
+	bmap->b_nchildren_per_block =
+		NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(bmap));
 	return 0;
 }
 
 void nilfs_btree_init_gc(struct nilfs_bmap *bmap)
 {
 	bmap->b_ops = &nilfs_btree_ops_gc;
+	bmap->b_nchildren_per_block =
+		NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(bmap));
 }
diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h
index 43c8c5b541f..22c02e35b6e 100644
--- a/fs/nilfs2/btree.h
+++ b/fs/nilfs2/btree.h
@@ -31,14 +31,6 @@
 #include "bmap.h"
 
 /**
- * struct nilfs_btree - B-tree structure
- * @bt_bmap: bmap base structure
- */
-struct nilfs_btree {
-	struct nilfs_bmap bt_bmap;
-};
-
-/**
  * struct nilfs_btree_path - A path on which B-tree operations are executed
  * @bp_bh: buffer head of node block
  * @bp_sib_bh: buffer head of sibling node block
@@ -54,7 +46,7 @@ struct nilfs_btree_path {
 	union nilfs_bmap_ptr_req bp_oldreq;
 	union nilfs_bmap_ptr_req bp_newreq;
 	struct nilfs_btnode_chkey_ctxt bp_ctxt;
-	void (*bp_op)(struct nilfs_btree *, struct nilfs_btree_path *,
+	void (*bp_op)(struct nilfs_bmap *, struct nilfs_btree_path *,
 		      int, __u64 *, __u64 *);
 };
 
@@ -80,4 +72,6 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *, __u64, __u64,
 				   const __u64 *, const __u64 *, int);
 void nilfs_btree_init_gc(struct nilfs_bmap *);
 
+int nilfs_btree_broken_node_block(struct buffer_head *bh);
+
 #endif	/* _NILFS_BTREE_H */
diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index 85c89dfc71f..b60277b4446 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -141,7 +141,7 @@ static void nilfs_check_page(struct page *page)
 	}
 	for (offs = 0; offs <= limit - NILFS_DIR_REC_LEN(1); offs += rec_len) {
 		p = (struct nilfs_dir_entry *)(kaddr + offs);
-		rec_len = le16_to_cpu(p->rec_len);
+		rec_len = nilfs_rec_len_from_disk(p->rec_len);
 
 		if (rec_len < NILFS_DIR_REC_LEN(1))
 			goto Eshort;
@@ -199,13 +199,10 @@ fail:
 static struct page *nilfs_get_page(struct inode *dir, unsigned long n)
 {
 	struct address_space *mapping = dir->i_mapping;
-	struct page *page = read_cache_page(mapping, n,
-				(filler_t *)mapping->a_ops->readpage, NULL);
+	struct page *page = read_mapping_page(mapping, n, NULL);
+
 	if (!IS_ERR(page)) {
-		wait_on_page_locked(page);
 		kmap(page);
-		if (!PageUptodate(page))
-			goto fail;
 		if (!PageChecked(page))
 			nilfs_check_page(page);
 		if (PageError(page))
@@ -238,7 +235,8 @@ nilfs_match(int len, const unsigned char *name, struct nilfs_dir_entry *de)
  */
 static struct nilfs_dir_entry *nilfs_next_entry(struct nilfs_dir_entry *p)
 {
-	return (struct nilfs_dir_entry *)((char *)p + le16_to_cpu(p->rec_len));
+	return (struct nilfs_dir_entry *)((char *)p +
+					  nilfs_rec_len_from_disk(p->rec_len));
 }
 
 static unsigned char
@@ -329,7 +327,7 @@ static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 					goto success;
 				}
 			}
-			filp->f_pos += le16_to_cpu(de->rec_len);
+			filp->f_pos += nilfs_rec_len_from_disk(de->rec_len);
 		}
 		nilfs_put_page(page);
 	}
@@ -444,7 +442,7 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de,
 		    struct page *page, struct inode *inode)
 {
 	unsigned from = (char *) de - (char *) page_address(page);
-	unsigned to = from + le16_to_cpu(de->rec_len);
+	unsigned to = from + nilfs_rec_len_from_disk(de->rec_len);
 	struct address_space *mapping = page->mapping;
 	int err;
 
@@ -500,7 +498,7 @@ int nilfs_add_link(struct dentry *dentry, struct inode *inode)
 				/* We hit i_size */
 				name_len = 0;
 				rec_len = chunk_size;
-				de->rec_len = cpu_to_le16(chunk_size);
+				de->rec_len = nilfs_rec_len_to_disk(chunk_size);
 				de->inode = 0;
 				goto got_it;
 			}
@@ -514,7 +512,7 @@ int nilfs_add_link(struct dentry *dentry, struct inode *inode)
 			if (nilfs_match(namelen, name, de))
 				goto out_unlock;
 			name_len = NILFS_DIR_REC_LEN(de->name_len);
-			rec_len = le16_to_cpu(de->rec_len);
+			rec_len = nilfs_rec_len_from_disk(de->rec_len);
 			if (!de->inode && rec_len >= reclen)
 				goto got_it;
 			if (rec_len >= name_len + reclen)
@@ -537,8 +535,8 @@ got_it:
 		struct nilfs_dir_entry *de1;
 
 		de1 = (struct nilfs_dir_entry *)((char *)de + name_len);
-		de1->rec_len = cpu_to_le16(rec_len - name_len);
-		de->rec_len = cpu_to_le16(name_len);
+		de1->rec_len = nilfs_rec_len_to_disk(rec_len - name_len);
+		de->rec_len = nilfs_rec_len_to_disk(name_len);
 		de = de1;
 	}
 	de->name_len = namelen;
@@ -569,7 +567,8 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page)
 	struct inode *inode = mapping->host;
 	char *kaddr = page_address(page);
 	unsigned from = ((char *)dir - kaddr) & ~(nilfs_chunk_size(inode) - 1);
-	unsigned to = ((char *)dir - kaddr) + le16_to_cpu(dir->rec_len);
+	unsigned to = ((char *)dir - kaddr) +
+		nilfs_rec_len_from_disk(dir->rec_len);
 	struct nilfs_dir_entry *pde = NULL;
 	struct nilfs_dir_entry *de = (struct nilfs_dir_entry *)(kaddr + from);
 	int err;
@@ -590,7 +589,7 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page)
 	err = nilfs_prepare_chunk(page, mapping, from, to);
 	BUG_ON(err);
 	if (pde)
-		pde->rec_len = cpu_to_le16(to - from);
+		pde->rec_len = nilfs_rec_len_to_disk(to - from);
 	dir->inode = 0;
 	nilfs_commit_chunk(page, mapping, from, to);
 	inode->i_ctime = inode->i_mtime = CURRENT_TIME;
@@ -624,14 +623,14 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent)
 	memset(kaddr, 0, chunk_size);
 	de = (struct nilfs_dir_entry *)kaddr;
 	de->name_len = 1;
-	de->rec_len = cpu_to_le16(NILFS_DIR_REC_LEN(1));
+	de->rec_len = nilfs_rec_len_to_disk(NILFS_DIR_REC_LEN(1));
 	memcpy(de->name, ".\0\0", 4);
 	de->inode = cpu_to_le64(inode->i_ino);
 	nilfs_set_de_type(de, inode);
 
 	de = (struct nilfs_dir_entry *)(kaddr + NILFS_DIR_REC_LEN(1));
 	de->name_len = 2;
-	de->rec_len = cpu_to_le16(chunk_size - NILFS_DIR_REC_LEN(1));
+	de->rec_len = nilfs_rec_len_to_disk(chunk_size - NILFS_DIR_REC_LEN(1));
 	de->inode = cpu_to_le64(parent->i_ino);
 	memcpy(de->name, "..\0", 4);
 	nilfs_set_de_type(de, inode);
diff --git a/fs/nilfs2/direct.c b/fs/nilfs2/direct.c
index 236753df5cd..324d80c5751 100644
--- a/fs/nilfs2/direct.c
+++ b/fs/nilfs2/direct.c
@@ -27,47 +27,43 @@
 #include "alloc.h"
 #include "dat.h"
 
-static inline __le64 *nilfs_direct_dptrs(const struct nilfs_direct *direct)
+static inline __le64 *nilfs_direct_dptrs(const struct nilfs_bmap *direct)
 {
 	return (__le64 *)
-		((struct nilfs_direct_node *)direct->d_bmap.b_u.u_data + 1);
+		((struct nilfs_direct_node *)direct->b_u.u_data + 1);
 }
 
 static inline __u64
-nilfs_direct_get_ptr(const struct nilfs_direct *direct, __u64 key)
+nilfs_direct_get_ptr(const struct nilfs_bmap *direct, __u64 key)
 {
-	return nilfs_bmap_dptr_to_ptr(*(nilfs_direct_dptrs(direct) + key));
+	return le64_to_cpu(*(nilfs_direct_dptrs(direct) + key));
 }
 
-static inline void nilfs_direct_set_ptr(struct nilfs_direct *direct,
+static inline void nilfs_direct_set_ptr(struct nilfs_bmap *direct,
 					__u64 key, __u64 ptr)
 {
-	*(nilfs_direct_dptrs(direct) + key) = nilfs_bmap_ptr_to_dptr(ptr);
+	*(nilfs_direct_dptrs(direct) + key) = cpu_to_le64(ptr);
 }
 
-static int nilfs_direct_lookup(const struct nilfs_bmap *bmap,
+static int nilfs_direct_lookup(const struct nilfs_bmap *direct,
 			       __u64 key, int level, __u64 *ptrp)
 {
-	struct nilfs_direct *direct;
 	__u64 ptr;
 
-	direct = (struct nilfs_direct *)bmap;  /* XXX: use macro for level 1 */
 	if (key > NILFS_DIRECT_KEY_MAX || level != 1)
 		return -ENOENT;
 	ptr = nilfs_direct_get_ptr(direct, key);
 	if (ptr == NILFS_BMAP_INVALID_PTR)
 		return -ENOENT;
 
-	if (ptrp != NULL)
-		*ptrp = ptr;
+	*ptrp = ptr;
 	return 0;
 }
 
-static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap,
+static int nilfs_direct_lookup_contig(const struct nilfs_bmap *direct,
 				      __u64 key, __u64 *ptrp,
 				      unsigned maxblocks)
 {
-	struct nilfs_direct *direct = (struct nilfs_direct *)bmap;
 	struct inode *dat = NULL;
 	__u64 ptr, ptr2;
 	sector_t blocknr;
@@ -79,8 +75,8 @@ static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap,
 	if (ptr == NILFS_BMAP_INVALID_PTR)
 		return -ENOENT;
 
-	if (NILFS_BMAP_USE_VBN(bmap)) {
-		dat = nilfs_bmap_get_dat(bmap);
+	if (NILFS_BMAP_USE_VBN(direct)) {
+		dat = nilfs_bmap_get_dat(direct);
 		ret = nilfs_dat_translate(dat, ptr, &blocknr);
 		if (ret < 0)
 			return ret;
@@ -106,29 +102,21 @@ static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap,
 }
 
 static __u64
-nilfs_direct_find_target_v(const struct nilfs_direct *direct, __u64 key)
+nilfs_direct_find_target_v(const struct nilfs_bmap *direct, __u64 key)
 {
 	__u64 ptr;
 
-	ptr = nilfs_bmap_find_target_seq(&direct->d_bmap, key);
+	ptr = nilfs_bmap_find_target_seq(direct, key);
 	if (ptr != NILFS_BMAP_INVALID_PTR)
 		/* sequential access */
 		return ptr;
 	else
 		/* block group */
-		return nilfs_bmap_find_target_in_group(&direct->d_bmap);
-}
-
-static void nilfs_direct_set_target_v(struct nilfs_direct *direct,
-				      __u64 key, __u64 ptr)
-{
-	direct->d_bmap.b_last_allocated_key = key;
-	direct->d_bmap.b_last_allocated_ptr = ptr;
+		return nilfs_bmap_find_target_in_group(direct);
 }
 
 static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
 {
-	struct nilfs_direct *direct = (struct nilfs_direct *)bmap;
 	union nilfs_bmap_ptr_req req;
 	struct inode *dat = NULL;
 	struct buffer_head *bh;
@@ -136,11 +124,11 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
 
 	if (key > NILFS_DIRECT_KEY_MAX)
 		return -ENOENT;
-	if (nilfs_direct_get_ptr(direct, key) != NILFS_BMAP_INVALID_PTR)
+	if (nilfs_direct_get_ptr(bmap, key) != NILFS_BMAP_INVALID_PTR)
 		return -EEXIST;
 
 	if (NILFS_BMAP_USE_VBN(bmap)) {
-		req.bpr_ptr = nilfs_direct_find_target_v(direct, key);
+		req.bpr_ptr = nilfs_direct_find_target_v(bmap, key);
 		dat = nilfs_bmap_get_dat(bmap);
 	}
 	ret = nilfs_bmap_prepare_alloc_ptr(bmap, &req, dat);
@@ -150,13 +138,13 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
 		set_buffer_nilfs_volatile(bh);
 
 		nilfs_bmap_commit_alloc_ptr(bmap, &req, dat);
-		nilfs_direct_set_ptr(direct, key, req.bpr_ptr);
+		nilfs_direct_set_ptr(bmap, key, req.bpr_ptr);
 
 		if (!nilfs_bmap_dirty(bmap))
 			nilfs_bmap_set_dirty(bmap);
 
 		if (NILFS_BMAP_USE_VBN(bmap))
-			nilfs_direct_set_target_v(direct, key, req.bpr_ptr);
+			nilfs_bmap_set_target_v(bmap, key, req.bpr_ptr);
 
 		nilfs_bmap_add_blocks(bmap, 1);
 	}
@@ -165,33 +153,30 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
 
 static int nilfs_direct_delete(struct nilfs_bmap *bmap, __u64 key)
 {
-	struct nilfs_direct *direct = (struct nilfs_direct *)bmap;
 	union nilfs_bmap_ptr_req req;
 	struct inode *dat;
 	int ret;
 
 	if (key > NILFS_DIRECT_KEY_MAX ||
-	    nilfs_direct_get_ptr(direct, key) == NILFS_BMAP_INVALID_PTR)
+	    nilfs_direct_get_ptr(bmap, key) == NILFS_BMAP_INVALID_PTR)
 		return -ENOENT;
 
 	dat = NILFS_BMAP_USE_VBN(bmap) ? nilfs_bmap_get_dat(bmap) : NULL;
-	req.bpr_ptr = nilfs_direct_get_ptr(direct, key);
+	req.bpr_ptr = nilfs_direct_get_ptr(bmap, key);
 
 	ret = nilfs_bmap_prepare_end_ptr(bmap, &req, dat);
 	if (!ret) {
 		nilfs_bmap_commit_end_ptr(bmap, &req, dat);
-		nilfs_direct_set_ptr(direct, key, NILFS_BMAP_INVALID_PTR);
+		nilfs_direct_set_ptr(bmap, key, NILFS_BMAP_INVALID_PTR);
 		nilfs_bmap_sub_blocks(bmap, 1);
 	}
 	return ret;
 }
 
-static int nilfs_direct_last_key(const struct nilfs_bmap *bmap, __u64 *keyp)
+static int nilfs_direct_last_key(const struct nilfs_bmap *direct, __u64 *keyp)
 {
-	struct nilfs_direct *direct;
 	__u64 key, lastkey;
 
-	direct = (struct nilfs_direct *)bmap;
 	lastkey = NILFS_DIRECT_KEY_MAX + 1;
 	for (key = NILFS_DIRECT_KEY_MIN; key <= NILFS_DIRECT_KEY_MAX; key++)
 		if (nilfs_direct_get_ptr(direct, key) !=
@@ -211,15 +196,13 @@ static int nilfs_direct_check_insert(const struct nilfs_bmap *bmap, __u64 key)
 	return key > NILFS_DIRECT_KEY_MAX;
 }
 
-static int nilfs_direct_gather_data(struct nilfs_bmap *bmap,
+static int nilfs_direct_gather_data(struct nilfs_bmap *direct,
 				    __u64 *keys, __u64 *ptrs, int nitems)
 {
-	struct nilfs_direct *direct;
 	__u64 key;
 	__u64 ptr;
 	int n;
 
-	direct = (struct nilfs_direct *)bmap;
 	if (nitems > NILFS_DIRECT_NBLOCKS)
 		nitems = NILFS_DIRECT_NBLOCKS;
 	n = 0;
@@ -237,7 +220,6 @@ static int nilfs_direct_gather_data(struct nilfs_bmap *bmap,
 int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap,
 				    __u64 key, __u64 *keys, __u64 *ptrs, int n)
 {
-	struct nilfs_direct *direct;
 	__le64 *dptrs;
 	int ret, i, j;
 
@@ -253,12 +235,11 @@ int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap,
 		bmap->b_ops->bop_clear(bmap);
 
 	/* convert */
-	direct = (struct nilfs_direct *)bmap;
-	dptrs = nilfs_direct_dptrs(direct);
+	dptrs = nilfs_direct_dptrs(bmap);
 	for (i = 0, j = 0; i < NILFS_DIRECT_NBLOCKS; i++) {
 		if ((j < n) && (i == keys[j])) {
 			dptrs[i] = (i != key) ?
-				nilfs_bmap_ptr_to_dptr(ptrs[j]) :
+				cpu_to_le64(ptrs[j]) :
 				NILFS_BMAP_INVALID_PTR;
 			j++;
 		} else
@@ -269,10 +250,9 @@ int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap,
 	return 0;
 }
 
-static int nilfs_direct_propagate(const struct nilfs_bmap *bmap,
+static int nilfs_direct_propagate(struct nilfs_bmap *bmap,
 				  struct buffer_head *bh)
 {
-	struct nilfs_direct *direct = (struct nilfs_direct *)bmap;
 	struct nilfs_palloc_req oldreq, newreq;
 	struct inode *dat;
 	__u64 key;
@@ -284,7 +264,7 @@ static int nilfs_direct_propagate(const struct nilfs_bmap *bmap,
 
 	dat = nilfs_bmap_get_dat(bmap);
 	key = nilfs_bmap_data_get_key(bmap, bh);
-	ptr = nilfs_direct_get_ptr(direct, key);
+	ptr = nilfs_direct_get_ptr(bmap, key);
 	if (!buffer_nilfs_volatile(bh)) {
 		oldreq.pr_entry_nr = ptr;
 		newreq.pr_entry_nr = ptr;
@@ -294,20 +274,20 @@ static int nilfs_direct_propagate(const struct nilfs_bmap *bmap,
 		nilfs_dat_commit_update(dat, &oldreq, &newreq,
 					bmap->b_ptr_type == NILFS_BMAP_PTR_VS);
 		set_buffer_nilfs_volatile(bh);
-		nilfs_direct_set_ptr(direct, key, newreq.pr_entry_nr);
+		nilfs_direct_set_ptr(bmap, key, newreq.pr_entry_nr);
 	} else
 		ret = nilfs_dat_mark_dirty(dat, ptr);
 
 	return ret;
 }
 
-static int nilfs_direct_assign_v(struct nilfs_direct *direct,
+static int nilfs_direct_assign_v(struct nilfs_bmap *direct,
 				 __u64 key, __u64 ptr,
 				 struct buffer_head **bh,
 				 sector_t blocknr,
 				 union nilfs_binfo *binfo)
 {
-	struct inode *dat = nilfs_bmap_get_dat(&direct->d_bmap);
+	struct inode *dat = nilfs_bmap_get_dat(direct);
 	union nilfs_bmap_ptr_req req;
 	int ret;
 
@@ -315,13 +295,13 @@ static int nilfs_direct_assign_v(struct nilfs_direct *direct,
 	ret = nilfs_dat_prepare_start(dat, &req.bpr_req);
 	if (!ret) {
 		nilfs_dat_commit_start(dat, &req.bpr_req, blocknr);
-		binfo->bi_v.bi_vblocknr = nilfs_bmap_ptr_to_dptr(ptr);
-		binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key);
+		binfo->bi_v.bi_vblocknr = cpu_to_le64(ptr);
+		binfo->bi_v.bi_blkoff = cpu_to_le64(key);
 	}
 	return ret;
 }
 
-static int nilfs_direct_assign_p(struct nilfs_direct *direct,
+static int nilfs_direct_assign_p(struct nilfs_bmap *direct,
 				 __u64 key, __u64 ptr,
 				 struct buffer_head **bh,
 				 sector_t blocknr,
@@ -329,7 +309,7 @@ static int nilfs_direct_assign_p(struct nilfs_direct *direct,
 {
 	nilfs_direct_set_ptr(direct, key, blocknr);
 
-	binfo->bi_dat.bi_blkoff = nilfs_bmap_key_to_dkey(key);
+	binfo->bi_dat.bi_blkoff = cpu_to_le64(key);
 	binfo->bi_dat.bi_level = 0;
 
 	return 0;
@@ -340,18 +320,16 @@ static int nilfs_direct_assign(struct nilfs_bmap *bmap,
 			       sector_t blocknr,
 			       union nilfs_binfo *binfo)
 {
-	struct nilfs_direct *direct;
 	__u64 key;
 	__u64 ptr;
 
-	direct = (struct nilfs_direct *)bmap;
 	key = nilfs_bmap_data_get_key(bmap, *bh);
 	if (unlikely(key > NILFS_DIRECT_KEY_MAX)) {
 		printk(KERN_CRIT "%s: invalid key: %llu\n", __func__,
 		       (unsigned long long)key);
 		return -EINVAL;
 	}
-	ptr = nilfs_direct_get_ptr(direct, key);
+	ptr = nilfs_direct_get_ptr(bmap, key);
 	if (unlikely(ptr == NILFS_BMAP_INVALID_PTR)) {
 		printk(KERN_CRIT "%s: invalid pointer: %llu\n", __func__,
 		       (unsigned long long)ptr);
@@ -359,8 +337,8 @@ static int nilfs_direct_assign(struct nilfs_bmap *bmap,
 	}
 
 	return NILFS_BMAP_USE_VBN(bmap) ?
-		nilfs_direct_assign_v(direct, key, ptr, bh, blocknr, binfo) :
-		nilfs_direct_assign_p(direct, key, ptr, bh, blocknr, binfo);
+		nilfs_direct_assign_v(bmap, key, ptr, bh, blocknr, binfo) :
+		nilfs_direct_assign_p(bmap, key, ptr, bh, blocknr, binfo);
 }
 
 static const struct nilfs_bmap_operations nilfs_direct_ops = {
diff --git a/fs/nilfs2/direct.h b/fs/nilfs2/direct.h
index a5ffd66e25d..dc643de20a2 100644
--- a/fs/nilfs2/direct.h
+++ b/fs/nilfs2/direct.h
@@ -28,8 +28,6 @@
 #include "bmap.h"
 
 
-struct nilfs_direct;
-
 /**
  * struct nilfs_direct_node - direct node
  * @dn_flags: flags
@@ -40,15 +38,6 @@ struct nilfs_direct_node {
 	__u8 pad[7];
 };
 
-/**
- * struct nilfs_direct - direct mapping
- * @d_bmap: bmap structure
- */
-struct nilfs_direct {
-	struct nilfs_bmap d_bmap;
-};
-
-
 #define NILFS_DIRECT_NBLOCKS	(NILFS_BMAP_SIZE / sizeof(__le64) - 1)
 #define NILFS_DIRECT_KEY_MIN	0
 #define NILFS_DIRECT_KEY_MAX	(NILFS_DIRECT_NBLOCKS - 1)
diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c
index 145f03cd7d3..bed3a783129 100644
--- a/fs/nilfs2/gcinode.c
+++ b/fs/nilfs2/gcinode.c
@@ -48,6 +48,8 @@
 #include <linux/slab.h>
 #include <linux/swap.h>
 #include "nilfs.h"
+#include "btree.h"
+#include "btnode.h"
 #include "page.h"
 #include "mdt.h"
 #include "dat.h"
@@ -149,8 +151,10 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff,
 int nilfs_gccache_submit_read_node(struct inode *inode, sector_t pbn,
 				   __u64 vbn, struct buffer_head **out_bh)
 {
-	int ret = nilfs_btnode_submit_block(&NILFS_I(inode)->i_btnode_cache,
-					    vbn ? : pbn, pbn, out_bh);
+	int ret;
+
+	ret = nilfs_btnode_submit_block(&NILFS_I(inode)->i_btnode_cache,
+					vbn ? : pbn, pbn, READ, out_bh, &pbn);
 	if (ret == -EEXIST) /* internal code (cache hit) */
 		ret = 0;
 	return ret;
@@ -164,10 +168,15 @@ int nilfs_gccache_wait_and_mark_dirty(struct buffer_head *bh)
 	if (buffer_dirty(bh))
 		return -EEXIST;
 
-	if (buffer_nilfs_node(bh))
+	if (buffer_nilfs_node(bh)) {
+		if (nilfs_btree_broken_node_block(bh)) {
+			clear_buffer_uptodate(bh);
+			return -EIO;
+		}
 		nilfs_btnode_mark_dirty(bh);
-	else
+	} else {
 		nilfs_mdt_mark_buffer_dirty(bh);
+	}
 	return 0;
 }
 
diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c
index 024be8c35bb..d01aff4957d 100644
--- a/fs/nilfs2/mdt.c
+++ b/fs/nilfs2/mdt.c
@@ -28,6 +28,7 @@
 #include <linux/swap.h>
 #include <linux/slab.h>
 #include "nilfs.h"
+#include "btnode.h"
 #include "segment.h"
 #include "page.h"
 #include "mdt.h"
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index 47d6d792812..0842d775b3e 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -32,7 +32,6 @@
 #include "the_nilfs.h"
 #include "sb.h"
 #include "bmap.h"
-#include "bmap_union.h"
 
 /*
  * nilfs inode data in memory
@@ -41,7 +40,7 @@ struct nilfs_inode_info {
 	__u32 i_flags;
 	unsigned long  i_state;		/* Dynamic state flags */
 	struct nilfs_bmap *i_bmap;
-	union nilfs_bmap_union i_bmap_union;
+	struct nilfs_bmap i_bmap_data;
 	__u64 i_xattr;	/* sector_t ??? */
 	__u32 i_dir_start_lookup;
 	__u64 i_cno;		/* check point number for GC inode */
@@ -71,9 +70,7 @@ static inline struct nilfs_inode_info *NILFS_I(const struct inode *inode)
 static inline struct nilfs_inode_info *
 NILFS_BMAP_I(const struct nilfs_bmap *bmap)
 {
-	return container_of((union nilfs_bmap_union *)bmap,
-			    struct nilfs_inode_info,
-			    i_bmap_union);
+	return container_of(bmap, struct nilfs_inode_info, i_bmap_data);
 }
 
 static inline struct inode *NILFS_BTNC_I(struct address_space *btnc)
@@ -107,6 +104,14 @@ enum {
 };
 
 /*
+ * commit flags for nilfs_commit_super and nilfs_sync_super
+ */
+enum {
+	NILFS_SB_COMMIT = 0,	/* Commit a super block alternately */
+	NILFS_SB_COMMIT_ALL	/* Commit both super blocks */
+};
+
+/*
  * Macros to check inode numbers
  */
 #define NILFS_MDT_INO_BITS   \
@@ -270,7 +275,14 @@ extern struct nilfs_super_block *
 nilfs_read_super_block(struct super_block *, u64, int, struct buffer_head **);
 extern int nilfs_store_magic_and_option(struct super_block *,
 					struct nilfs_super_block *, char *);
+extern int nilfs_check_feature_compatibility(struct super_block *,
+					     struct nilfs_super_block *);
+extern void nilfs_set_log_cursor(struct nilfs_super_block *,
+				 struct the_nilfs *);
+extern struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *,
+						      int flip);
 extern int nilfs_commit_super(struct nilfs_sb_info *, int);
+extern int nilfs_cleanup_super(struct nilfs_sb_info *);
 extern int nilfs_attach_checkpoint(struct nilfs_sb_info *, __u64);
 extern void nilfs_detach_checkpoint(struct nilfs_sb_info *);
 
diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c
index 8de3e1e4813..aab11db2cb0 100644
--- a/fs/nilfs2/page.c
+++ b/fs/nilfs2/page.c
@@ -37,7 +37,8 @@
 
 #define NILFS_BUFFER_INHERENT_BITS  \
 	((1UL << BH_Uptodate) | (1UL << BH_Mapped) | (1UL << BH_NILFS_Node) | \
-	 (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated))
+	 (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated) | \
+	 (1UL << BH_NILFS_Checked))
 
 static struct buffer_head *
 __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index,
@@ -129,6 +130,7 @@ void nilfs_forget_buffer(struct buffer_head *bh)
 
 	lock_buffer(bh);
 	clear_buffer_nilfs_volatile(bh);
+	clear_buffer_nilfs_checked(bh);
 	clear_buffer_dirty(bh);
 	if (nilfs_page_buffers_clean(page))
 		__nilfs_clear_page_dirty(page);
@@ -480,6 +482,7 @@ void nilfs_clear_dirty_pages(struct address_space *mapping)
 				lock_buffer(bh);
 				clear_buffer_dirty(bh);
 				clear_buffer_nilfs_volatile(bh);
+				clear_buffer_nilfs_checked(bh);
 				clear_buffer_uptodate(bh);
 				clear_buffer_mapped(bh);
 				unlock_buffer(bh);
diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h
index 8abca4d1c1f..f53d8da41ed 100644
--- a/fs/nilfs2/page.h
+++ b/fs/nilfs2/page.h
@@ -34,11 +34,13 @@ enum {
 	BH_NILFS_Allocated = BH_PrivateStart,
 	BH_NILFS_Node,
 	BH_NILFS_Volatile,
+	BH_NILFS_Checked,
 };
 
 BUFFER_FNS(NILFS_Allocated, nilfs_allocated)	/* nilfs private buffers */
 BUFFER_FNS(NILFS_Node, nilfs_node)		/* nilfs node buffers */
 BUFFER_FNS(NILFS_Volatile, nilfs_volatile)
+BUFFER_FNS(NILFS_Checked, nilfs_checked)	/* buffer is verified */
 
 
 void nilfs_mark_buffer_dirty(struct buffer_head *bh);
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index bae2a516b4e..83e3d8c61a0 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -91,27 +91,9 @@ static int nilfs_warn_segment_error(int err)
 	return -EINVAL;
 }
 
-static void store_segsum_info(struct nilfs_segsum_info *ssi,
-			      struct nilfs_segment_summary *sum,
-			      unsigned int blocksize)
-{
-	ssi->flags = le16_to_cpu(sum->ss_flags);
-	ssi->seg_seq = le64_to_cpu(sum->ss_seq);
-	ssi->ctime = le64_to_cpu(sum->ss_create);
-	ssi->next = le64_to_cpu(sum->ss_next);
-	ssi->nblocks = le32_to_cpu(sum->ss_nblocks);
-	ssi->nfinfo = le32_to_cpu(sum->ss_nfinfo);
-	ssi->sumbytes = le32_to_cpu(sum->ss_sumbytes);
-
-	ssi->nsumblk = DIV_ROUND_UP(ssi->sumbytes, blocksize);
-	ssi->nfileblk = ssi->nblocks - ssi->nsumblk - !!NILFS_SEG_HAS_SR(ssi);
-
-	/* need to verify ->ss_bytes field if read ->ss_cno */
-}
-
 /**
- * calc_crc_cont - check CRC of blocks continuously
- * @sbi: nilfs_sb_info
+ * nilfs_compute_checksum - compute checksum of blocks continuously
+ * @nilfs: nilfs object
  * @bhs: buffer head of start block
  * @sum: place to store result
  * @offset: offset bytes in the first block
@@ -119,23 +101,25 @@ static void store_segsum_info(struct nilfs_segsum_info *ssi,
  * @start: DBN of start block
  * @nblock: number of blocks to be checked
  */
-static int calc_crc_cont(struct nilfs_sb_info *sbi, struct buffer_head *bhs,
-			 u32 *sum, unsigned long offset, u64 check_bytes,
-			 sector_t start, unsigned long nblock)
+static int nilfs_compute_checksum(struct the_nilfs *nilfs,
+				  struct buffer_head *bhs, u32 *sum,
+				  unsigned long offset, u64 check_bytes,
+				  sector_t start, unsigned long nblock)
 {
-	unsigned long blocksize = sbi->s_super->s_blocksize;
+	unsigned int blocksize = nilfs->ns_blocksize;
 	unsigned long size;
 	u32 crc;
 
 	BUG_ON(offset >= blocksize);
 	check_bytes -= offset;
 	size = min_t(u64, check_bytes, blocksize - offset);
-	crc = crc32_le(sbi->s_nilfs->ns_crc_seed,
+	crc = crc32_le(nilfs->ns_crc_seed,
 		       (unsigned char *)bhs->b_data + offset, size);
 	if (--nblock > 0) {
 		do {
-			struct buffer_head *bh
-				= sb_bread(sbi->s_super, ++start);
+			struct buffer_head *bh;
+
+			bh = __bread(nilfs->ns_bdev, ++start, blocksize);
 			if (!bh)
 				return -EIO;
 			check_bytes -= size;
@@ -150,12 +134,12 @@ static int calc_crc_cont(struct nilfs_sb_info *sbi, struct buffer_head *bhs,
 
 /**
  * nilfs_read_super_root_block - read super root block
- * @sb: super_block
+ * @nilfs: nilfs object
  * @sr_block: disk block number of the super root block
  * @pbh: address of a buffer_head pointer to return super root buffer
  * @check: CRC check flag
  */
-int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block,
+int nilfs_read_super_root_block(struct the_nilfs *nilfs, sector_t sr_block,
 				struct buffer_head **pbh, int check)
 {
 	struct buffer_head *bh_sr;
@@ -164,7 +148,7 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block,
 	int ret;
 
 	*pbh = NULL;
-	bh_sr = sb_bread(sb, sr_block);
+	bh_sr = __bread(nilfs->ns_bdev, sr_block, nilfs->ns_blocksize);
 	if (unlikely(!bh_sr)) {
 		ret = NILFS_SEG_FAIL_IO;
 		goto failed;
@@ -174,12 +158,13 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block,
 	if (check) {
 		unsigned bytes = le16_to_cpu(sr->sr_bytes);
 
-		if (bytes == 0 || bytes > sb->s_blocksize) {
+		if (bytes == 0 || bytes > nilfs->ns_blocksize) {
 			ret = NILFS_SEG_FAIL_CHECKSUM_SUPER_ROOT;
 			goto failed_bh;
 		}
-		if (calc_crc_cont(NILFS_SB(sb), bh_sr, &crc,
-				  sizeof(sr->sr_sum), bytes, sr_block, 1)) {
+		if (nilfs_compute_checksum(
+			    nilfs, bh_sr, &crc, sizeof(sr->sr_sum), bytes,
+			    sr_block, 1)) {
 			ret = NILFS_SEG_FAIL_IO;
 			goto failed_bh;
 		}
@@ -199,64 +184,76 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block,
 }
 
 /**
- * load_segment_summary - read segment summary of the specified partial segment
- * @sbi: nilfs_sb_info
- * @pseg_start: start disk block number of partial segment
- * @seg_seq: sequence number requested
- * @ssi: pointer to nilfs_segsum_info struct to store information
+ * nilfs_read_log_header - read summary header of the specified log
+ * @nilfs: nilfs object
+ * @start_blocknr: start block number of the log
+ * @sum: pointer to return segment summary structure
  */
-static int
-load_segment_summary(struct nilfs_sb_info *sbi, sector_t pseg_start,
-		     u64 seg_seq, struct nilfs_segsum_info *ssi)
+static struct buffer_head *
+nilfs_read_log_header(struct the_nilfs *nilfs, sector_t start_blocknr,
+		      struct nilfs_segment_summary **sum)
 {
 	struct buffer_head *bh_sum;
-	struct nilfs_segment_summary *sum;
+
+	bh_sum = __bread(nilfs->ns_bdev, start_blocknr, nilfs->ns_blocksize);
+	if (bh_sum)
+		*sum = (struct nilfs_segment_summary *)bh_sum->b_data;
+	return bh_sum;
+}
+
+/**
+ * nilfs_validate_log - verify consistency of log
+ * @nilfs: nilfs object
+ * @seg_seq: sequence number of segment
+ * @bh_sum: buffer head of summary block
+ * @sum: segment summary struct
+ */
+static int nilfs_validate_log(struct the_nilfs *nilfs, u64 seg_seq,
+			      struct buffer_head *bh_sum,
+			      struct nilfs_segment_summary *sum)
+{
 	unsigned long nblock;
 	u32 crc;
-	int ret = NILFS_SEG_FAIL_IO;
+	int ret;
 
-	bh_sum = sb_bread(sbi->s_super, pseg_start);
-	if (!bh_sum)
+	ret = NILFS_SEG_FAIL_MAGIC;
+	if (le32_to_cpu(sum->ss_magic) != NILFS_SEGSUM_MAGIC)
 		goto out;
 
-	sum = (struct nilfs_segment_summary *)bh_sum->b_data;
-
-	/* Check consistency of segment summary */
-	if (le32_to_cpu(sum->ss_magic) != NILFS_SEGSUM_MAGIC) {
-		ret = NILFS_SEG_FAIL_MAGIC;
-		goto failed;
-	}
-	store_segsum_info(ssi, sum, sbi->s_super->s_blocksize);
-	if (seg_seq != ssi->seg_seq) {
-		ret = NILFS_SEG_FAIL_SEQ;
-		goto failed;
-	}
+	ret = NILFS_SEG_FAIL_SEQ;
+	if (le64_to_cpu(sum->ss_seq) != seg_seq)
+		goto out;
 
-	nblock = ssi->nblocks;
-	if (unlikely(nblock == 0 ||
-		     nblock > sbi->s_nilfs->ns_blocks_per_segment)) {
+	nblock = le32_to_cpu(sum->ss_nblocks);
+	ret = NILFS_SEG_FAIL_CONSISTENCY;
+	if (unlikely(nblock == 0 || nblock > nilfs->ns_blocks_per_segment))
 		/* This limits the number of blocks read in the CRC check */
-		ret = NILFS_SEG_FAIL_CONSISTENCY;
-		goto failed;
-	}
-	if (calc_crc_cont(sbi, bh_sum, &crc, sizeof(sum->ss_datasum),
-			  ((u64)nblock << sbi->s_super->s_blocksize_bits),
-			  pseg_start, nblock)) {
-		ret = NILFS_SEG_FAIL_IO;
-		goto failed;
-	}
-	if (crc == le32_to_cpu(sum->ss_datasum))
-		ret = 0;
-	else
-		ret = NILFS_SEG_FAIL_CHECKSUM_FULL;
- failed:
-	brelse(bh_sum);
- out:
+		goto out;
+
+	ret = NILFS_SEG_FAIL_IO;
+	if (nilfs_compute_checksum(nilfs, bh_sum, &crc, sizeof(sum->ss_datasum),
+				   ((u64)nblock << nilfs->ns_blocksize_bits),
+				   bh_sum->b_blocknr, nblock))
+		goto out;
+
+	ret = NILFS_SEG_FAIL_CHECKSUM_FULL;
+	if (crc != le32_to_cpu(sum->ss_datasum))
+		goto out;
+	ret = 0;
+out:
 	return ret;
 }
 
-static void *segsum_get(struct super_block *sb, struct buffer_head **pbh,
-			unsigned int *offset, unsigned int bytes)
+/**
+ * nilfs_read_summary_info - read an item on summary blocks of a log
+ * @nilfs: nilfs object
+ * @pbh: the current buffer head on summary blocks [in, out]
+ * @offset: the current byte offset on summary blocks [in, out]
+ * @bytes: byte size of the item to be read
+ */
+static void *nilfs_read_summary_info(struct the_nilfs *nilfs,
+				     struct buffer_head **pbh,
+				     unsigned int *offset, unsigned int bytes)
 {
 	void *ptr;
 	sector_t blocknr;
@@ -265,7 +262,8 @@ static void *segsum_get(struct super_block *sb, struct buffer_head **pbh,
 	if (bytes > (*pbh)->b_size - *offset) {
 		blocknr = (*pbh)->b_blocknr;
 		brelse(*pbh);
-		*pbh = sb_bread(sb, blocknr + 1);
+		*pbh = __bread(nilfs->ns_bdev, blocknr + 1,
+			       nilfs->ns_blocksize);
 		if (unlikely(!*pbh))
 			return NULL;
 		*offset = 0;
@@ -275,9 +273,18 @@ static void *segsum_get(struct super_block *sb, struct buffer_head **pbh,
 	return ptr;
 }
 
-static void segsum_skip(struct super_block *sb, struct buffer_head **pbh,
-			unsigned int *offset, unsigned int bytes,
-			unsigned long count)
+/**
+ * nilfs_skip_summary_info - skip items on summary blocks of a log
+ * @nilfs: nilfs object
+ * @pbh: the current buffer head on summary blocks [in, out]
+ * @offset: the current byte offset on summary blocks [in, out]
+ * @bytes: byte size of the item to be skipped
+ * @count: number of items to be skipped
+ */
+static void nilfs_skip_summary_info(struct the_nilfs *nilfs,
+				    struct buffer_head **pbh,
+				    unsigned int *offset, unsigned int bytes,
+				    unsigned long count)
 {
 	unsigned int rest_item_in_current_block
 		= ((*pbh)->b_size - *offset) / bytes;
@@ -294,36 +301,46 @@ static void segsum_skip(struct super_block *sb, struct buffer_head **pbh,
 		*offset = bytes * (count - (bcnt - 1) * nitem_per_block);
 
 		brelse(*pbh);
-		*pbh = sb_bread(sb, blocknr + bcnt);
+		*pbh = __bread(nilfs->ns_bdev, blocknr + bcnt,
+			       nilfs->ns_blocksize);
 	}
 }
 
-static int
-collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr,
-			   struct nilfs_segsum_info *ssi,
-			   struct list_head *head)
+/**
+ * nilfs_scan_dsync_log - get block information of a log written for data sync
+ * @nilfs: nilfs object
+ * @start_blocknr: start block number of the log
+ * @sum: log summary information
+ * @head: list head to add nilfs_recovery_block struct
+ */
+static int nilfs_scan_dsync_log(struct the_nilfs *nilfs, sector_t start_blocknr,
+				struct nilfs_segment_summary *sum,
+				struct list_head *head)
 {
 	struct buffer_head *bh;
 	unsigned int offset;
-	unsigned long nfinfo = ssi->nfinfo;
-	sector_t blocknr = sum_blocknr + ssi->nsumblk;
+	u32 nfinfo, sumbytes;
+	sector_t blocknr;
 	ino_t ino;
 	int err = -EIO;
 
+	nfinfo = le32_to_cpu(sum->ss_nfinfo);
 	if (!nfinfo)
 		return 0;
 
-	bh = sb_bread(sbi->s_super, sum_blocknr);
+	sumbytes = le32_to_cpu(sum->ss_sumbytes);
+	blocknr = start_blocknr + DIV_ROUND_UP(sumbytes, nilfs->ns_blocksize);
+	bh = __bread(nilfs->ns_bdev, start_blocknr, nilfs->ns_blocksize);
 	if (unlikely(!bh))
 		goto out;
 
-	offset = le16_to_cpu(
-		((struct nilfs_segment_summary *)bh->b_data)->ss_bytes);
+	offset = le16_to_cpu(sum->ss_bytes);
 	for (;;) {
 		unsigned long nblocks, ndatablk, nnodeblk;
 		struct nilfs_finfo *finfo;
 
-		finfo = segsum_get(sbi->s_super, &bh, &offset, sizeof(*finfo));
+		finfo = nilfs_read_summary_info(nilfs, &bh, &offset,
+						sizeof(*finfo));
 		if (unlikely(!finfo))
 			goto out;
 
@@ -336,8 +353,8 @@ collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr,
 			struct nilfs_recovery_block *rb;
 			struct nilfs_binfo_v *binfo;
 
-			binfo = segsum_get(sbi->s_super, &bh, &offset,
-					   sizeof(*binfo));
+			binfo = nilfs_read_summary_info(nilfs, &bh, &offset,
+							sizeof(*binfo));
 			if (unlikely(!binfo))
 				goto out;
 
@@ -355,9 +372,9 @@ collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr,
 		}
 		if (--nfinfo == 0)
 			break;
-		blocknr += nnodeblk; /* always 0 for the data sync segments */
-		segsum_skip(sbi->s_super, &bh, &offset, sizeof(__le64),
-			    nnodeblk);
+		blocknr += nnodeblk; /* always 0 for data sync logs */
+		nilfs_skip_summary_info(nilfs, &bh, &offset, sizeof(__le64),
+					nnodeblk);
 		if (unlikely(!bh))
 			goto out;
 	}
@@ -467,14 +484,14 @@ static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs,
 	return err;
 }
 
-static int nilfs_recovery_copy_block(struct nilfs_sb_info *sbi,
+static int nilfs_recovery_copy_block(struct the_nilfs *nilfs,
 				     struct nilfs_recovery_block *rb,
 				     struct page *page)
 {
 	struct buffer_head *bh_org;
 	void *kaddr;
 
-	bh_org = sb_bread(sbi->s_super, rb->blocknr);
+	bh_org = __bread(nilfs->ns_bdev, rb->blocknr, nilfs->ns_blocksize);
 	if (unlikely(!bh_org))
 		return -EIO;
 
@@ -485,13 +502,14 @@ static int nilfs_recovery_copy_block(struct nilfs_sb_info *sbi,
 	return 0;
 }
 
-static int recover_dsync_blocks(struct nilfs_sb_info *sbi,
-				struct list_head *head,
-				unsigned long *nr_salvaged_blocks)
+static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs,
+				      struct nilfs_sb_info *sbi,
+				      struct list_head *head,
+				      unsigned long *nr_salvaged_blocks)
 {
 	struct inode *inode;
 	struct nilfs_recovery_block *rb, *n;
-	unsigned blocksize = sbi->s_super->s_blocksize;
+	unsigned blocksize = nilfs->ns_blocksize;
 	struct page *page;
 	loff_t pos;
 	int err = 0, err2 = 0;
@@ -511,7 +529,7 @@ static int recover_dsync_blocks(struct nilfs_sb_info *sbi,
 		if (unlikely(err))
 			goto failed_inode;
 
-		err = nilfs_recovery_copy_block(sbi, rb, page);
+		err = nilfs_recovery_copy_block(nilfs, rb, page);
 		if (unlikely(err))
 			goto failed_page;
 
@@ -551,18 +569,20 @@ static int recover_dsync_blocks(struct nilfs_sb_info *sbi,
 /**
  * nilfs_do_roll_forward - salvage logical segments newer than the latest
  * checkpoint
+ * @nilfs: nilfs object
  * @sbi: nilfs_sb_info
- * @nilfs: the_nilfs
  * @ri: pointer to a nilfs_recovery_info
  */
 static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 				 struct nilfs_sb_info *sbi,
 				 struct nilfs_recovery_info *ri)
 {
-	struct nilfs_segsum_info ssi;
+	struct buffer_head *bh_sum = NULL;
+	struct nilfs_segment_summary *sum;
 	sector_t pseg_start;
 	sector_t seg_start, seg_end;  /* Starting/ending DBN of full segment */
 	unsigned long nsalvaged_blocks = 0;
+	unsigned int flags;
 	u64 seg_seq;
 	__u64 segnum, nextnum = 0;
 	int empty_seg = 0;
@@ -581,8 +601,14 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 	nilfs_get_segment_range(nilfs, segnum, &seg_start, &seg_end);
 
 	while (segnum != ri->ri_segnum || pseg_start <= ri->ri_pseg_start) {
+		brelse(bh_sum);
+		bh_sum = nilfs_read_log_header(nilfs, pseg_start, &sum);
+		if (!bh_sum) {
+			err = -EIO;
+			goto failed;
+		}
 
-		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi);
+		ret = nilfs_validate_log(nilfs, seg_seq, bh_sum, sum);
 		if (ret) {
 			if (ret == NILFS_SEG_FAIL_IO) {
 				err = -EIO;
@@ -590,33 +616,38 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 			}
 			goto strayed;
 		}
-		if (unlikely(NILFS_SEG_HAS_SR(&ssi)))
+
+		flags = le16_to_cpu(sum->ss_flags);
+		if (flags & NILFS_SS_SR)
 			goto confused;
 
 		/* Found a valid partial segment; do recovery actions */
-		nextnum = nilfs_get_segnum_of_block(nilfs, ssi.next);
+		nextnum = nilfs_get_segnum_of_block(nilfs,
+						    le64_to_cpu(sum->ss_next));
 		empty_seg = 0;
-		nilfs->ns_ctime = ssi.ctime;
-		if (!(ssi.flags & NILFS_SS_GC))
-			nilfs->ns_nongc_ctime = ssi.ctime;
+		nilfs->ns_ctime = le64_to_cpu(sum->ss_create);
+		if (!(flags & NILFS_SS_GC))
+			nilfs->ns_nongc_ctime = nilfs->ns_ctime;
 
 		switch (state) {
 		case RF_INIT_ST:
-			if (!NILFS_SEG_LOGBGN(&ssi) || !NILFS_SEG_DSYNC(&ssi))
+			if (!(flags & NILFS_SS_LOGBGN) ||
+			    !(flags & NILFS_SS_SYNDT))
 				goto try_next_pseg;
 			state = RF_DSYNC_ST;
 			/* Fall through */
 		case RF_DSYNC_ST:
-			if (!NILFS_SEG_DSYNC(&ssi))
+			if (!(flags & NILFS_SS_SYNDT))
 				goto confused;
 
-			err = collect_blocks_from_segsum(
-				sbi, pseg_start, &ssi, &dsync_blocks);
+			err = nilfs_scan_dsync_log(nilfs, pseg_start, sum,
+						   &dsync_blocks);
 			if (unlikely(err))
 				goto failed;
-			if (NILFS_SEG_LOGEND(&ssi)) {
-				err = recover_dsync_blocks(
-					sbi, &dsync_blocks, &nsalvaged_blocks);
+			if (flags & NILFS_SS_LOGEND) {
+				err = nilfs_recover_dsync_blocks(
+					nilfs, sbi, &dsync_blocks,
+					&nsalvaged_blocks);
 				if (unlikely(err))
 					goto failed;
 				state = RF_INIT_ST;
@@ -627,7 +658,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
  try_next_pseg:
 		if (pseg_start == ri->ri_lsegs_end)
 			break;
-		pseg_start += ssi.nblocks;
+		pseg_start += le32_to_cpu(sum->ss_nblocks);
 		if (pseg_start < seg_end)
 			continue;
 		goto feed_segment;
@@ -652,8 +683,9 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 		ri->ri_need_recovery = NILFS_RECOVERY_ROLLFORWARD_DONE;
 	}
  out:
+	brelse(bh_sum);
 	dispose_recovery_list(&dsync_blocks);
-	nilfs_detach_writer(sbi->s_nilfs, sbi);
+	nilfs_detach_writer(nilfs, sbi);
 	return err;
 
  confused:
@@ -667,7 +699,6 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 }
 
 static void nilfs_finish_roll_forward(struct the_nilfs *nilfs,
-				      struct nilfs_sb_info *sbi,
 				      struct nilfs_recovery_info *ri)
 {
 	struct buffer_head *bh;
@@ -677,7 +708,7 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs,
 	    nilfs_get_segnum_of_block(nilfs, ri->ri_super_root))
 		return;
 
-	bh = sb_getblk(sbi->s_super, ri->ri_lsegs_start);
+	bh = __getblk(nilfs->ns_bdev, ri->ri_lsegs_start, nilfs->ns_blocksize);
 	BUG_ON(!bh);
 	memset(bh->b_data, 0, bh->b_size);
 	set_buffer_dirty(bh);
@@ -690,9 +721,8 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs,
 }
 
 /**
- * nilfs_recover_logical_segments - salvage logical segments written after
- * the latest super root
- * @nilfs: the_nilfs
+ * nilfs_salvage_orphan_logs - salvage logs written after the latest checkpoint
+ * @nilfs: nilfs object
  * @sbi: nilfs_sb_info
  * @ri: pointer to a nilfs_recovery_info struct to store search results.
  *
@@ -709,9 +739,9 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs,
  *
  * %-ENOMEM - Insufficient memory available.
  */
-int nilfs_recover_logical_segments(struct the_nilfs *nilfs,
-				   struct nilfs_sb_info *sbi,
-				   struct nilfs_recovery_info *ri)
+int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs,
+			      struct nilfs_sb_info *sbi,
+			      struct nilfs_recovery_info *ri)
 {
 	int err;
 
@@ -751,7 +781,7 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs,
 			goto failed;
 		}
 
-		nilfs_finish_roll_forward(nilfs, sbi, ri);
+		nilfs_finish_roll_forward(nilfs, ri);
 	}
 
  failed:
@@ -762,7 +792,6 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs,
 /**
  * nilfs_search_super_root - search the latest valid super root
  * @nilfs: the_nilfs
- * @sbi: nilfs_sb_info
  * @ri: pointer to a nilfs_recovery_info struct to store search results.
  *
  * nilfs_search_super_root() looks for the latest super-root from a partial
@@ -775,14 +804,19 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs,
  * %-EINVAL - No valid segment found
  *
  * %-EIO - I/O error
+ *
+ * %-ENOMEM - Insufficient memory available.
  */
-int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
+int nilfs_search_super_root(struct the_nilfs *nilfs,
 			    struct nilfs_recovery_info *ri)
 {
-	struct nilfs_segsum_info ssi;
+	struct buffer_head *bh_sum = NULL;
+	struct nilfs_segment_summary *sum;
 	sector_t pseg_start, pseg_end, sr_pseg_start = 0;
 	sector_t seg_start, seg_end; /* range of full segment (block number) */
 	sector_t b, end;
+	unsigned long nblocks;
+	unsigned int flags;
 	u64 seg_seq;
 	__u64 segnum, nextnum = 0;
 	__u64 cno;
@@ -801,17 +835,24 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 	/* Read ahead segment */
 	b = seg_start;
 	while (b <= seg_end)
-		sb_breadahead(sbi->s_super, b++);
+		__breadahead(nilfs->ns_bdev, b++, nilfs->ns_blocksize);
 
 	for (;;) {
-		/* Load segment summary */
-		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi);
+		brelse(bh_sum);
+		ret = NILFS_SEG_FAIL_IO;
+		bh_sum = nilfs_read_log_header(nilfs, pseg_start, &sum);
+		if (!bh_sum)
+			goto failed;
+
+		ret = nilfs_validate_log(nilfs, seg_seq, bh_sum, sum);
 		if (ret) {
 			if (ret == NILFS_SEG_FAIL_IO)
 				goto failed;
 			goto strayed;
 		}
-		pseg_end = pseg_start + ssi.nblocks - 1;
+
+		nblocks = le32_to_cpu(sum->ss_nblocks);
+		pseg_end = pseg_start + nblocks - 1;
 		if (unlikely(pseg_end > seg_end)) {
 			ret = NILFS_SEG_FAIL_CONSISTENCY;
 			goto strayed;
@@ -821,11 +862,13 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 		ri->ri_pseg_start = pseg_start;
 		ri->ri_seq = seg_seq;
 		ri->ri_segnum = segnum;
-		nextnum = nilfs_get_segnum_of_block(nilfs, ssi.next);
+		nextnum = nilfs_get_segnum_of_block(nilfs,
+						    le64_to_cpu(sum->ss_next));
 		ri->ri_nextnum = nextnum;
 		empty_seg = 0;
 
-		if (!NILFS_SEG_HAS_SR(&ssi) && !scan_newer) {
+		flags = le16_to_cpu(sum->ss_flags);
+		if (!(flags & NILFS_SS_SR) && !scan_newer) {
 			/* This will never happen because a superblock
 			   (last_segment) always points to a pseg
 			   having a super root. */
@@ -836,14 +879,15 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 		if (pseg_start == seg_start) {
 			nilfs_get_segment_range(nilfs, nextnum, &b, &end);
 			while (b <= end)
-				sb_breadahead(sbi->s_super, b++);
+				__breadahead(nilfs->ns_bdev, b++,
+					     nilfs->ns_blocksize);
 		}
-		if (!NILFS_SEG_HAS_SR(&ssi)) {
-			if (!ri->ri_lsegs_start && NILFS_SEG_LOGBGN(&ssi)) {
+		if (!(flags & NILFS_SS_SR)) {
+			if (!ri->ri_lsegs_start && (flags & NILFS_SS_LOGBGN)) {
 				ri->ri_lsegs_start = pseg_start;
 				ri->ri_lsegs_start_seq = seg_seq;
 			}
-			if (NILFS_SEG_LOGEND(&ssi))
+			if (flags & NILFS_SS_LOGEND)
 				ri->ri_lsegs_end = pseg_start;
 			goto try_next_pseg;
 		}
@@ -854,12 +898,12 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 		ri->ri_lsegs_start = ri->ri_lsegs_end = 0;
 
 		nilfs_dispose_segment_list(&segments);
-		nilfs->ns_pseg_offset = (sr_pseg_start = pseg_start)
-			+ ssi.nblocks - seg_start;
+		sr_pseg_start = pseg_start;
+		nilfs->ns_pseg_offset = pseg_start + nblocks - seg_start;
 		nilfs->ns_seg_seq = seg_seq;
 		nilfs->ns_segnum = segnum;
 		nilfs->ns_cno = cno;  /* nilfs->ns_cno = ri->ri_cno + 1 */
-		nilfs->ns_ctime = ssi.ctime;
+		nilfs->ns_ctime = le64_to_cpu(sum->ss_create);
 		nilfs->ns_nextnum = nextnum;
 
 		if (scan_newer)
@@ -870,15 +914,9 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 			scan_newer = 1;
 		}
 
-		/* reset region for roll-forward */
-		pseg_start += ssi.nblocks;
-		if (pseg_start < seg_end)
-			continue;
-		goto feed_segment;
-
  try_next_pseg:
 		/* Standing on a course, or met an inconsistent state */
-		pseg_start += ssi.nblocks;
+		pseg_start += nblocks;
 		if (pseg_start < seg_end)
 			continue;
 		goto feed_segment;
@@ -909,6 +947,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 
  super_root_found:
 	/* Updating pointers relating to the latest checkpoint */
+	brelse(bh_sum);
 	list_splice_tail(&segments, &ri->ri_used_segments);
 	nilfs->ns_last_pseg = sr_pseg_start;
 	nilfs->ns_last_seq = nilfs->ns_seg_seq;
@@ -916,6 +955,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 	return 0;
 
  failed:
+	brelse(bh_sum);
 	nilfs_dispose_segment_list(&segments);
 	return (ret < 0) ? ret : nilfs_warn_segment_error(ret);
 }
diff --git a/fs/nilfs2/segbuf.h b/fs/nilfs2/segbuf.h
index 85fbb66455e..b04f08cc239 100644
--- a/fs/nilfs2/segbuf.h
+++ b/fs/nilfs2/segbuf.h
@@ -54,17 +54,6 @@ struct nilfs_segsum_info {
 	sector_t		next;
 };
 
-/* macro for the flags */
-#define NILFS_SEG_HAS_SR(sum)    ((sum)->flags & NILFS_SS_SR)
-#define NILFS_SEG_LOGBGN(sum)    ((sum)->flags & NILFS_SS_LOGBGN)
-#define NILFS_SEG_LOGEND(sum)    ((sum)->flags & NILFS_SS_LOGEND)
-#define NILFS_SEG_DSYNC(sum)     ((sum)->flags & NILFS_SS_SYNDT)
-#define NILFS_SEG_SIMPLEX(sum) \
-	(((sum)->flags & (NILFS_SS_LOGBGN | NILFS_SS_LOGEND)) == \
-	 (NILFS_SS_LOGBGN | NILFS_SS_LOGEND))
-
-#define NILFS_SEG_EMPTY(sum)	((sum)->nblocks == (sum)->nsumblk)
-
 /**
  * struct nilfs_segment_buffer - Segment buffer
  * @sb_super: back pointer to a superblock struct
@@ -141,6 +130,19 @@ int nilfs_segbuf_extend_payload(struct nilfs_segment_buffer *,
 				struct buffer_head **);
 void nilfs_segbuf_fill_in_segsum(struct nilfs_segment_buffer *);
 
+static inline int nilfs_segbuf_simplex(struct nilfs_segment_buffer *segbuf)
+{
+	unsigned int flags = segbuf->sb_sum.flags;
+
+	return (flags & (NILFS_SS_LOGBGN | NILFS_SS_LOGEND)) ==
+		(NILFS_SS_LOGBGN | NILFS_SS_LOGEND);
+}
+
+static inline int nilfs_segbuf_empty(struct nilfs_segment_buffer *segbuf)
+{
+	return segbuf->sb_sum.nblocks == segbuf->sb_sum.nsumblk;
+}
+
 static inline void
 nilfs_segbuf_add_segsum_buffer(struct nilfs_segment_buffer *segbuf,
 			       struct buffer_head *bh)
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index c9201649cc4..9fd051a33c4 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -1914,12 +1914,12 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci)
 			}
 		}
 
-		if (!NILFS_SEG_SIMPLEX(&segbuf->sb_sum)) {
-			if (NILFS_SEG_LOGBGN(&segbuf->sb_sum)) {
+		if (!nilfs_segbuf_simplex(segbuf)) {
+			if (segbuf->sb_sum.flags & NILFS_SS_LOGBGN) {
 				set_bit(NILFS_SC_UNCLOSED, &sci->sc_flags);
 				sci->sc_lseg_stime = jiffies;
 			}
-			if (NILFS_SEG_LOGEND(&segbuf->sb_sum))
+			if (segbuf->sb_sum.flags & NILFS_SS_LOGEND)
 				clear_bit(NILFS_SC_UNCLOSED, &sci->sc_flags);
 		}
 	}
@@ -1951,7 +1951,6 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci)
 	if (update_sr) {
 		nilfs_set_last_segment(nilfs, segbuf->sb_pseg_start,
 				       segbuf->sb_sum.seg_seq, nilfs->ns_cno++);
-		set_nilfs_sb_dirty(nilfs);
 
 		clear_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags);
 		clear_bit(NILFS_SC_DIRTY, &sci->sc_flags);
@@ -2082,7 +2081,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode)
 
 		/* Avoid empty segment */
 		if (sci->sc_stage.scnt == NILFS_ST_DONE &&
-		    NILFS_SEG_EMPTY(&sci->sc_curseg->sb_sum)) {
+		    nilfs_segbuf_empty(sci->sc_curseg)) {
 			nilfs_segctor_abort_construction(sci, nilfs, 1);
 			goto out;
 		}
@@ -2408,6 +2407,7 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode)
 {
 	struct nilfs_sb_info *sbi = sci->sc_sbi;
 	struct the_nilfs *nilfs = sbi->s_nilfs;
+	struct nilfs_super_block **sbp;
 	int err = 0;
 
 	nilfs_segctor_accept(sci);
@@ -2423,8 +2423,13 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode)
 		if (test_bit(NILFS_SC_SUPER_ROOT, &sci->sc_flags) &&
 		    nilfs_discontinued(nilfs)) {
 			down_write(&nilfs->ns_sem);
-			err = nilfs_commit_super(
-				sbi, nilfs_altsb_need_update(nilfs));
+			err = -EIO;
+			sbp = nilfs_prepare_super(sbi,
+						  nilfs_sb_will_flip(nilfs));
+			if (likely(sbp)) {
+				nilfs_set_log_cursor(sbp[0], nilfs);
+				err = nilfs_commit_super(sbi, NILFS_SB_COMMIT);
+			}
 			up_write(&nilfs->ns_sem);
 		}
 	}
diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h
index 01e20dbb217..17c487bd815 100644
--- a/fs/nilfs2/segment.h
+++ b/fs/nilfs2/segment.h
@@ -234,13 +234,13 @@ extern int nilfs_attach_segment_constructor(struct nilfs_sb_info *);
 extern void nilfs_detach_segment_constructor(struct nilfs_sb_info *);
 
 /* recovery.c */
-extern int nilfs_read_super_root_block(struct super_block *, sector_t,
+extern int nilfs_read_super_root_block(struct the_nilfs *, sector_t,
 				       struct buffer_head **, int);
-extern int nilfs_search_super_root(struct the_nilfs *, struct nilfs_sb_info *,
+extern int nilfs_search_super_root(struct the_nilfs *,
 				   struct nilfs_recovery_info *);
-extern int nilfs_recover_logical_segments(struct the_nilfs *,
-					  struct nilfs_sb_info *,
-					  struct nilfs_recovery_info *);
+extern int nilfs_salvage_orphan_logs(struct the_nilfs *,
+				     struct nilfs_sb_info *,
+				     struct nilfs_recovery_info *);
 extern void nilfs_dispose_segment_list(struct list_head *);
 
 #endif /* _NILFS_SEGMENT_H */
diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
index 414ef68931c..26078b3407c 100644
--- a/fs/nilfs2/super.c
+++ b/fs/nilfs2/super.c
@@ -55,6 +55,8 @@
 #include "nilfs.h"
 #include "mdt.h"
 #include "alloc.h"
+#include "btree.h"
+#include "btnode.h"
 #include "page.h"
 #include "cpfile.h"
 #include "ifile.h"
@@ -74,6 +76,25 @@ struct kmem_cache *nilfs_btree_path_cache;
 
 static int nilfs_remount(struct super_block *sb, int *flags, char *data);
 
+static void nilfs_set_error(struct nilfs_sb_info *sbi)
+{
+	struct the_nilfs *nilfs = sbi->s_nilfs;
+	struct nilfs_super_block **sbp;
+
+	down_write(&nilfs->ns_sem);
+	if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) {
+		nilfs->ns_mount_state |= NILFS_ERROR_FS;
+		sbp = nilfs_prepare_super(sbi, 0);
+		if (likely(sbp)) {
+			sbp[0]->s_state |= cpu_to_le16(NILFS_ERROR_FS);
+			if (sbp[1])
+				sbp[1]->s_state |= cpu_to_le16(NILFS_ERROR_FS);
+			nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);
+		}
+	}
+	up_write(&nilfs->ns_sem);
+}
+
 /**
  * nilfs_error() - report failure condition on a filesystem
  *
@@ -99,16 +120,7 @@ void nilfs_error(struct super_block *sb, const char *function,
 	va_end(args);
 
 	if (!(sb->s_flags & MS_RDONLY)) {
-		struct the_nilfs *nilfs = sbi->s_nilfs;
-
-		down_write(&nilfs->ns_sem);
-		if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) {
-			nilfs->ns_mount_state |= NILFS_ERROR_FS;
-			nilfs->ns_sbp[0]->s_state |=
-				cpu_to_le16(NILFS_ERROR_FS);
-			nilfs_commit_super(sbi, 1);
-		}
-		up_write(&nilfs->ns_sem);
+		nilfs_set_error(sbi);
 
 		if (nilfs_test_opt(sbi, ERRORS_RO)) {
 			printk(KERN_CRIT "Remounting filesystem read-only\n");
@@ -176,7 +188,7 @@ static void nilfs_clear_inode(struct inode *inode)
 	nilfs_btnode_cache_clear(&ii->i_btnode_cache);
 }
 
-static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb)
+static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag)
 {
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	int err;
@@ -202,12 +214,20 @@ static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb)
 		printk(KERN_ERR
 		       "NILFS: unable to write superblock (err=%d)\n", err);
 		if (err == -EIO && nilfs->ns_sbh[1]) {
+			/*
+			 * sbp[0] points to newer log than sbp[1],
+			 * so copy sbp[0] to sbp[1] to take over sbp[0].
+			 */
+			memcpy(nilfs->ns_sbp[1], nilfs->ns_sbp[0],
+			       nilfs->ns_sbsize);
 			nilfs_fall_back_super_block(nilfs);
 			goto retry;
 		}
 	} else {
 		struct nilfs_super_block *sbp = nilfs->ns_sbp[0];
 
+		nilfs->ns_sbwcount++;
+
 		/*
 		 * The latest segment becomes trailable from the position
 		 * written in superblock.
@@ -216,66 +236,122 @@ static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb)
 
 		/* update GC protection for recent segments */
 		if (nilfs->ns_sbh[1]) {
-			sbp = NULL;
-			if (dupsb) {
+			if (flag == NILFS_SB_COMMIT_ALL) {
 				set_buffer_dirty(nilfs->ns_sbh[1]);
-				if (!sync_dirty_buffer(nilfs->ns_sbh[1]))
-					sbp = nilfs->ns_sbp[1];
+				if (sync_dirty_buffer(nilfs->ns_sbh[1]) < 0)
+					goto out;
 			}
+			if (le64_to_cpu(nilfs->ns_sbp[1]->s_last_cno) <
+			    le64_to_cpu(nilfs->ns_sbp[0]->s_last_cno))
+				sbp = nilfs->ns_sbp[1];
 		}
-		if (sbp) {
-			spin_lock(&nilfs->ns_last_segment_lock);
-			nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq);
-			spin_unlock(&nilfs->ns_last_segment_lock);
-		}
-	}
 
+		spin_lock(&nilfs->ns_last_segment_lock);
+		nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq);
+		spin_unlock(&nilfs->ns_last_segment_lock);
+	}
+ out:
 	return err;
 }
 
-int nilfs_commit_super(struct nilfs_sb_info *sbi, int dupsb)
+void nilfs_set_log_cursor(struct nilfs_super_block *sbp,
+			  struct the_nilfs *nilfs)
+{
+	sector_t nfreeblocks;
+
+	/* nilfs->ns_sem must be locked by the caller. */
+	nilfs_count_free_blocks(nilfs, &nfreeblocks);
+	sbp->s_free_blocks_count = cpu_to_le64(nfreeblocks);
+
+	spin_lock(&nilfs->ns_last_segment_lock);
+	sbp->s_last_seq = cpu_to_le64(nilfs->ns_last_seq);
+	sbp->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg);
+	sbp->s_last_cno = cpu_to_le64(nilfs->ns_last_cno);
+	spin_unlock(&nilfs->ns_last_segment_lock);
+}
+
+struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *sbi,
+					       int flip)
 {
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	struct nilfs_super_block **sbp = nilfs->ns_sbp;
-	sector_t nfreeblocks;
-	time_t t;
-	int err;
 
-	/* nilfs->sem must be locked by the caller. */
+	/* nilfs->ns_sem must be locked by the caller. */
 	if (sbp[0]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {
-		if (sbp[1] && sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC))
-			nilfs_swap_super_block(nilfs);
-		else {
+		if (sbp[1] &&
+		    sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) {
+			memcpy(sbp[0], sbp[1], nilfs->ns_sbsize);
+		} else {
 			printk(KERN_CRIT "NILFS: superblock broke on dev %s\n",
 			       sbi->s_super->s_id);
-			return -EIO;
+			return NULL;
 		}
+	} else if (sbp[1] &&
+		   sbp[1]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {
+			memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
 	}
-	err = nilfs_count_free_blocks(nilfs, &nfreeblocks);
-	if (unlikely(err)) {
-		printk(KERN_ERR "NILFS: failed to count free blocks\n");
-		return err;
-	}
-	spin_lock(&nilfs->ns_last_segment_lock);
-	sbp[0]->s_last_seq = cpu_to_le64(nilfs->ns_last_seq);
-	sbp[0]->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg);
-	sbp[0]->s_last_cno = cpu_to_le64(nilfs->ns_last_cno);
-	spin_unlock(&nilfs->ns_last_segment_lock);
 
+	if (flip && sbp[1])
+		nilfs_swap_super_block(nilfs);
+
+	return sbp;
+}
+
+int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag)
+{
+	struct the_nilfs *nilfs = sbi->s_nilfs;
+	struct nilfs_super_block **sbp = nilfs->ns_sbp;
+	time_t t;
+
+	/* nilfs->ns_sem must be locked by the caller. */
 	t = get_seconds();
-	nilfs->ns_sbwtime[0] = t;
-	sbp[0]->s_free_blocks_count = cpu_to_le64(nfreeblocks);
+	nilfs->ns_sbwtime = t;
 	sbp[0]->s_wtime = cpu_to_le64(t);
 	sbp[0]->s_sum = 0;
 	sbp[0]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,
 					     (unsigned char *)sbp[0],
 					     nilfs->ns_sbsize));
-	if (dupsb && sbp[1]) {
-		memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
-		nilfs->ns_sbwtime[1] = t;
+	if (flag == NILFS_SB_COMMIT_ALL && sbp[1]) {
+		sbp[1]->s_wtime = sbp[0]->s_wtime;
+		sbp[1]->s_sum = 0;
+		sbp[1]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,
+					    (unsigned char *)sbp[1],
+					    nilfs->ns_sbsize));
 	}
 	clear_nilfs_sb_dirty(nilfs);
-	return nilfs_sync_super(sbi, dupsb);
+	return nilfs_sync_super(sbi, flag);
+}
+
+/**
+ * nilfs_cleanup_super() - write filesystem state for cleanup
+ * @sbi: nilfs_sb_info to be unmounted or degraded to read-only
+ *
+ * This function restores state flags in the on-disk super block.
+ * This will set "clean" flag (i.e. NILFS_VALID_FS) unless the
+ * filesystem was not clean previously.
+ */
+int nilfs_cleanup_super(struct nilfs_sb_info *sbi)
+{
+	struct nilfs_super_block **sbp;
+	int flag = NILFS_SB_COMMIT;
+	int ret = -EIO;
+
+	sbp = nilfs_prepare_super(sbi, 0);
+	if (sbp) {
+		sbp[0]->s_state = cpu_to_le16(sbi->s_nilfs->ns_mount_state);
+		nilfs_set_log_cursor(sbp[0], sbi->s_nilfs);
+		if (sbp[1] && sbp[0]->s_last_cno == sbp[1]->s_last_cno) {
+			/*
+			 * make the "clean" flag also to the opposite
+			 * super block if both super blocks point to
+			 * the same checkpoint.
+			 */
+			sbp[1]->s_state = sbp[0]->s_state;
+			flag = NILFS_SB_COMMIT_ALL;
+		}
+		ret = nilfs_commit_super(sbi, flag);
+	}
+	return ret;
 }
 
 static void nilfs_put_super(struct super_block *sb)
@@ -289,8 +365,7 @@ static void nilfs_put_super(struct super_block *sb)
 
 	if (!(sb->s_flags & MS_RDONLY)) {
 		down_write(&nilfs->ns_sem);
-		nilfs->ns_sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state);
-		nilfs_commit_super(sbi, 1);
+		nilfs_cleanup_super(sbi);
 		up_write(&nilfs->ns_sem);
 	}
 	down_write(&nilfs->ns_super_sem);
@@ -311,6 +386,7 @@ static int nilfs_sync_fs(struct super_block *sb, int wait)
 {
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
 	struct the_nilfs *nilfs = sbi->s_nilfs;
+	struct nilfs_super_block **sbp;
 	int err = 0;
 
 	/* This function is called when super block should be written back */
@@ -318,8 +394,13 @@ static int nilfs_sync_fs(struct super_block *sb, int wait)
 		err = nilfs_construct_segment(sb);
 
 	down_write(&nilfs->ns_sem);
-	if (nilfs_sb_dirty(nilfs))
-		nilfs_commit_super(sbi, 1);
+	if (nilfs_sb_dirty(nilfs)) {
+		sbp = nilfs_prepare_super(sbi, nilfs_sb_will_flip(nilfs));
+		if (likely(sbp)) {
+			nilfs_set_log_cursor(sbp[0], nilfs);
+			nilfs_commit_super(sbi, NILFS_SB_COMMIT);
+		}
+	}
 	up_write(&nilfs->ns_sem);
 
 	return err;
@@ -442,20 +523,20 @@ static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
 
 	if (!nilfs_test_opt(sbi, BARRIER))
-		seq_printf(seq, ",nobarrier");
+		seq_puts(seq, ",nobarrier");
 	if (nilfs_test_opt(sbi, SNAPSHOT))
 		seq_printf(seq, ",cp=%llu",
 			   (unsigned long long int)sbi->s_snapshot_cno);
 	if (nilfs_test_opt(sbi, ERRORS_PANIC))
-		seq_printf(seq, ",errors=panic");
+		seq_puts(seq, ",errors=panic");
 	if (nilfs_test_opt(sbi, ERRORS_CONT))
-		seq_printf(seq, ",errors=continue");
+		seq_puts(seq, ",errors=continue");
 	if (nilfs_test_opt(sbi, STRICT_ORDER))
-		seq_printf(seq, ",order=strict");
+		seq_puts(seq, ",order=strict");
 	if (nilfs_test_opt(sbi, NORECOVERY))
-		seq_printf(seq, ",norecovery");
+		seq_puts(seq, ",norecovery");
 	if (nilfs_test_opt(sbi, DISCARD))
-		seq_printf(seq, ",discard");
+		seq_puts(seq, ",discard");
 
 	return 0;
 }
@@ -524,23 +605,25 @@ static const struct export_operations nilfs_export_ops = {
 
 enum {
 	Opt_err_cont, Opt_err_panic, Opt_err_ro,
-	Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery,
-	Opt_discard, Opt_err,
+	Opt_barrier, Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery,
+	Opt_discard, Opt_nodiscard, Opt_err,
 };
 
 static match_table_t tokens = {
 	{Opt_err_cont, "errors=continue"},
 	{Opt_err_panic, "errors=panic"},
 	{Opt_err_ro, "errors=remount-ro"},
+	{Opt_barrier, "barrier"},
 	{Opt_nobarrier, "nobarrier"},
 	{Opt_snapshot, "cp=%u"},
 	{Opt_order, "order=%s"},
 	{Opt_norecovery, "norecovery"},
 	{Opt_discard, "discard"},
+	{Opt_nodiscard, "nodiscard"},
 	{Opt_err, NULL}
 };
 
-static int parse_options(char *options, struct super_block *sb)
+static int parse_options(char *options, struct super_block *sb, int is_remount)
 {
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
 	char *p;
@@ -557,6 +640,9 @@ static int parse_options(char *options, struct super_block *sb)
 
 		token = match_token(p, tokens, args);
 		switch (token) {
+		case Opt_barrier:
+			nilfs_set_opt(sbi, BARRIER);
+			break;
 		case Opt_nobarrier:
 			nilfs_clear_opt(sbi, BARRIER);
 			break;
@@ -582,8 +668,26 @@ static int parse_options(char *options, struct super_block *sb)
 		case Opt_snapshot:
 			if (match_int(&args[0], &option) || option <= 0)
 				return 0;
-			if (!(sb->s_flags & MS_RDONLY))
+			if (is_remount) {
+				if (!nilfs_test_opt(sbi, SNAPSHOT)) {
+					printk(KERN_ERR
+					       "NILFS: cannot change regular "
+					       "mount to snapshot.\n");
+					return 0;
+				} else if (option != sbi->s_snapshot_cno) {
+					printk(KERN_ERR
+					       "NILFS: cannot remount to a "
+					       "different snapshot.\n");
+					return 0;
+				}
+				break;
+			}
+			if (!(sb->s_flags & MS_RDONLY)) {
+				printk(KERN_ERR "NILFS: cannot mount snapshot "
+				       "read/write.  A read-only option is "
+				       "required.\n");
 				return 0;
+			}
 			sbi->s_snapshot_cno = option;
 			nilfs_set_opt(sbi, SNAPSHOT);
 			break;
@@ -593,6 +697,9 @@ static int parse_options(char *options, struct super_block *sb)
 		case Opt_discard:
 			nilfs_set_opt(sbi, DISCARD);
 			break;
+		case Opt_nodiscard:
+			nilfs_clear_opt(sbi, DISCARD);
+			break;
 		default:
 			printk(KERN_ERR
 			       "NILFS: Unrecognized mount option \"%s\"\n", p);
@@ -613,11 +720,18 @@ nilfs_set_default_options(struct nilfs_sb_info *sbi,
 static int nilfs_setup_super(struct nilfs_sb_info *sbi)
 {
 	struct the_nilfs *nilfs = sbi->s_nilfs;
-	struct nilfs_super_block *sbp = nilfs->ns_sbp[0];
-	int max_mnt_count = le16_to_cpu(sbp->s_max_mnt_count);
-	int mnt_count = le16_to_cpu(sbp->s_mnt_count);
+	struct nilfs_super_block **sbp;
+	int max_mnt_count;
+	int mnt_count;
+
+	/* nilfs->ns_sem must be locked by the caller. */
+	sbp = nilfs_prepare_super(sbi, 0);
+	if (!sbp)
+		return -EIO;
+
+	max_mnt_count = le16_to_cpu(sbp[0]->s_max_mnt_count);
+	mnt_count = le16_to_cpu(sbp[0]->s_mnt_count);
 
-	/* nilfs->sem must be locked by the caller. */
 	if (nilfs->ns_mount_state & NILFS_ERROR_FS) {
 		printk(KERN_WARNING
 		       "NILFS warning: mounting fs with errors\n");
@@ -628,12 +742,15 @@ static int nilfs_setup_super(struct nilfs_sb_info *sbi)
 #endif
 	}
 	if (!max_mnt_count)
-		sbp->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT);
-
-	sbp->s_mnt_count = cpu_to_le16(mnt_count + 1);
-	sbp->s_state = cpu_to_le16(le16_to_cpu(sbp->s_state) & ~NILFS_VALID_FS);
-	sbp->s_mtime = cpu_to_le64(get_seconds());
-	return nilfs_commit_super(sbi, 1);
+		sbp[0]->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT);
+
+	sbp[0]->s_mnt_count = cpu_to_le16(mnt_count + 1);
+	sbp[0]->s_state =
+		cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & ~NILFS_VALID_FS);
+	sbp[0]->s_mtime = cpu_to_le64(get_seconds());
+	/* synchronize sbp[1] with sbp[0] */
+	memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
+	return nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);
 }
 
 struct nilfs_super_block *nilfs_read_super_block(struct super_block *sb,
@@ -670,7 +787,31 @@ int nilfs_store_magic_and_option(struct super_block *sb,
 	sbi->s_interval = le32_to_cpu(sbp->s_c_interval);
 	sbi->s_watermark = le32_to_cpu(sbp->s_c_block_max);
 
-	return !parse_options(data, sb) ? -EINVAL : 0 ;
+	return !parse_options(data, sb, 0) ? -EINVAL : 0 ;
+}
+
+int nilfs_check_feature_compatibility(struct super_block *sb,
+				      struct nilfs_super_block *sbp)
+{
+	__u64 features;
+
+	features = le64_to_cpu(sbp->s_feature_incompat) &
+		~NILFS_FEATURE_INCOMPAT_SUPP;
+	if (features) {
+		printk(KERN_ERR "NILFS: couldn't mount because of unsupported "
+		       "optional features (%llx)\n",
+		       (unsigned long long)features);
+		return -EINVAL;
+	}
+	features = le64_to_cpu(sbp->s_feature_compat_ro) &
+		~NILFS_FEATURE_COMPAT_RO_SUPP;
+	if (!(sb->s_flags & MS_RDONLY) && features) {
+		printk(KERN_ERR "NILFS: couldn't mount RDWR because of "
+		       "unsupported optional features (%llx)\n",
+		       (unsigned long long)features);
+		return -EINVAL;
+	}
+	return 0;
 }
 
 /**
@@ -819,7 +960,6 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent,
 static int nilfs_remount(struct super_block *sb, int *flags, char *data)
 {
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
-	struct nilfs_super_block *sbp;
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	unsigned long old_sb_flags;
 	struct nilfs_mount_options old_opts;
@@ -833,32 +973,17 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data)
 	old_opts.snapshot_cno = sbi->s_snapshot_cno;
 	was_snapshot = nilfs_test_opt(sbi, SNAPSHOT);
 
-	if (!parse_options(data, sb)) {
+	if (!parse_options(data, sb, 1)) {
 		err = -EINVAL;
 		goto restore_opts;
 	}
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL);
 
 	err = -EINVAL;
-	if (was_snapshot) {
-		if (!(*flags & MS_RDONLY)) {
-			printk(KERN_ERR "NILFS (device %s): cannot remount "
-			       "snapshot read/write.\n",
-			       sb->s_id);
-			goto restore_opts;
-		} else if (sbi->s_snapshot_cno != old_opts.snapshot_cno) {
-			printk(KERN_ERR "NILFS (device %s): cannot "
-			       "remount to a different snapshot.\n",
-			       sb->s_id);
-			goto restore_opts;
-		}
-	} else {
-		if (nilfs_test_opt(sbi, SNAPSHOT)) {
-			printk(KERN_ERR "NILFS (device %s): cannot change "
-			       "a regular mount to a snapshot.\n",
-			       sb->s_id);
-			goto restore_opts;
-		}
+	if (was_snapshot && !(*flags & MS_RDONLY)) {
+		printk(KERN_ERR "NILFS (device %s): cannot remount snapshot "
+		       "read/write.\n", sb->s_id);
+		goto restore_opts;
 	}
 
 	if (!nilfs_valid_fs(nilfs)) {
@@ -880,19 +1005,29 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data)
 		 * the RDONLY flag and then mark the partition as valid again.
 		 */
 		down_write(&nilfs->ns_sem);
-		sbp = nilfs->ns_sbp[0];
-		if (!(sbp->s_state & le16_to_cpu(NILFS_VALID_FS)) &&
-		    (nilfs->ns_mount_state & NILFS_VALID_FS))
-			sbp->s_state = cpu_to_le16(nilfs->ns_mount_state);
-		sbp->s_mtime = cpu_to_le64(get_seconds());
-		nilfs_commit_super(sbi, 1);
+		nilfs_cleanup_super(sbi);
 		up_write(&nilfs->ns_sem);
 	} else {
+		__u64 features;
+
 		/*
 		 * Mounting a RDONLY partition read-write, so reread and
 		 * store the current valid flag.  (It may have been changed
 		 * by fsck since we originally mounted the partition.)
 		 */
+		down_read(&nilfs->ns_sem);
+		features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) &
+			~NILFS_FEATURE_COMPAT_RO_SUPP;
+		up_read(&nilfs->ns_sem);
+		if (features) {
+			printk(KERN_WARNING "NILFS (device %s): couldn't "
+			       "remount RDWR because of unsupported optional "
+			       "features (%llx)\n",
+			       sb->s_id, (unsigned long long)features);
+			err = -EROFS;
+			goto restore_opts;
+		}
+
 		sb->s_flags &= ~MS_RDONLY;
 
 		err = nilfs_attach_segment_constructor(sbi);
@@ -1119,7 +1254,7 @@ static void nilfs_inode_init_once(void *obj)
 	init_rwsem(&ii->xattr_sem);
 #endif
 	nilfs_btnode_cache_init_once(&ii->i_btnode_cache);
-	ii->i_bmap = (struct nilfs_bmap *)&ii->i_bmap_union;
+	ii->i_bmap = &ii->i_bmap_data;
 	inode_init_once(&ii->vfs_inode);
 }
 
diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c
index 8c1097327ab..37de1f062d8 100644
--- a/fs/nilfs2/the_nilfs.c
+++ b/fs/nilfs2/the_nilfs.c
@@ -38,6 +38,8 @@
 static LIST_HEAD(nilfs_objects);
 static DEFINE_SPINLOCK(nilfs_lock);
 
+static int nilfs_valid_sb(struct nilfs_super_block *sbp);
+
 void nilfs_set_last_segment(struct the_nilfs *nilfs,
 			    sector_t start_blocknr, u64 seq, __u64 cno)
 {
@@ -45,6 +47,16 @@ void nilfs_set_last_segment(struct the_nilfs *nilfs,
 	nilfs->ns_last_pseg = start_blocknr;
 	nilfs->ns_last_seq = seq;
 	nilfs->ns_last_cno = cno;
+
+	if (!nilfs_sb_dirty(nilfs)) {
+		if (nilfs->ns_prev_seq == nilfs->ns_last_seq)
+			goto stay_cursor;
+
+		set_nilfs_sb_dirty(nilfs);
+	}
+	nilfs->ns_prev_seq = nilfs->ns_last_seq;
+
+ stay_cursor:
 	spin_unlock(&nilfs->ns_last_segment_lock);
 }
 
@@ -159,8 +171,7 @@ void put_nilfs(struct the_nilfs *nilfs)
 	kfree(nilfs);
 }
 
-static int nilfs_load_super_root(struct the_nilfs *nilfs,
-				 struct nilfs_sb_info *sbi, sector_t sr_block)
+static int nilfs_load_super_root(struct the_nilfs *nilfs, sector_t sr_block)
 {
 	struct buffer_head *bh_sr;
 	struct nilfs_super_root *raw_sr;
@@ -169,7 +180,7 @@ static int nilfs_load_super_root(struct the_nilfs *nilfs,
 	unsigned inode_size;
 	int err;
 
-	err = nilfs_read_super_root_block(sbi->s_super, sr_block, &bh_sr, 1);
+	err = nilfs_read_super_root_block(nilfs, sr_block, &bh_sr, 1);
 	if (unlikely(err))
 		return err;
 
@@ -248,6 +259,37 @@ static void nilfs_clear_recovery_info(struct nilfs_recovery_info *ri)
 }
 
 /**
+ * nilfs_store_log_cursor - load log cursor from a super block
+ * @nilfs: nilfs object
+ * @sbp: buffer storing super block to be read
+ *
+ * nilfs_store_log_cursor() reads the last position of the log
+ * containing a super root from a given super block, and initializes
+ * relevant information on the nilfs object preparatory for log
+ * scanning and recovery.
+ */
+static int nilfs_store_log_cursor(struct the_nilfs *nilfs,
+				  struct nilfs_super_block *sbp)
+{
+	int ret = 0;
+
+	nilfs->ns_last_pseg = le64_to_cpu(sbp->s_last_pseg);
+	nilfs->ns_last_cno = le64_to_cpu(sbp->s_last_cno);
+	nilfs->ns_last_seq = le64_to_cpu(sbp->s_last_seq);
+
+	nilfs->ns_prev_seq = nilfs->ns_last_seq;
+	nilfs->ns_seg_seq = nilfs->ns_last_seq;
+	nilfs->ns_segnum =
+		nilfs_get_segnum_of_block(nilfs, nilfs->ns_last_pseg);
+	nilfs->ns_cno = nilfs->ns_last_cno + 1;
+	if (nilfs->ns_segnum >= nilfs->ns_nsegments) {
+		printk(KERN_ERR "NILFS invalid last segment number.\n");
+		ret = -EINVAL;
+	}
+	return ret;
+}
+
+/**
  * load_nilfs - load and recover the nilfs
  * @nilfs: the_nilfs structure to be released
  * @sbi: nilfs_sb_info used to recover past segment
@@ -285,13 +327,55 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi)
 
 	nilfs_init_recovery_info(&ri);
 
-	err = nilfs_search_super_root(nilfs, sbi, &ri);
+	err = nilfs_search_super_root(nilfs, &ri);
 	if (unlikely(err)) {
-		printk(KERN_ERR "NILFS: error searching super root.\n");
-		goto failed;
+		struct nilfs_super_block **sbp = nilfs->ns_sbp;
+		int blocksize;
+
+		if (err != -EINVAL)
+			goto scan_error;
+
+		if (!nilfs_valid_sb(sbp[1])) {
+			printk(KERN_WARNING
+			       "NILFS warning: unable to fall back to spare"
+			       "super block\n");
+			goto scan_error;
+		}
+		printk(KERN_INFO
+		       "NILFS: try rollback from an earlier position\n");
+
+		/*
+		 * restore super block with its spare and reconfigure
+		 * relevant states of the nilfs object.
+		 */
+		memcpy(sbp[0], sbp[1], nilfs->ns_sbsize);
+		nilfs->ns_crc_seed = le32_to_cpu(sbp[0]->s_crc_seed);
+		nilfs->ns_sbwtime = le64_to_cpu(sbp[0]->s_wtime);
+
+		/* verify consistency between two super blocks */
+		blocksize = BLOCK_SIZE << le32_to_cpu(sbp[0]->s_log_block_size);
+		if (blocksize != nilfs->ns_blocksize) {
+			printk(KERN_WARNING
+			       "NILFS warning: blocksize differs between "
+			       "two super blocks (%d != %d)\n",
+			       blocksize, nilfs->ns_blocksize);
+			goto scan_error;
+		}
+
+		err = nilfs_store_log_cursor(nilfs, sbp[0]);
+		if (err)
+			goto scan_error;
+
+		/* drop clean flag to allow roll-forward and recovery */
+		nilfs->ns_mount_state &= ~NILFS_VALID_FS;
+		valid_fs = 0;
+
+		err = nilfs_search_super_root(nilfs, &ri);
+		if (err)
+			goto scan_error;
 	}
 
-	err = nilfs_load_super_root(nilfs, sbi, ri.ri_super_root);
+	err = nilfs_load_super_root(nilfs, ri.ri_super_root);
 	if (unlikely(err)) {
 		printk(KERN_ERR "NILFS: error loading super root.\n");
 		goto failed;
@@ -301,11 +385,23 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi)
 		goto skip_recovery;
 
 	if (s_flags & MS_RDONLY) {
+		__u64 features;
+
 		if (nilfs_test_opt(sbi, NORECOVERY)) {
 			printk(KERN_INFO "NILFS: norecovery option specified. "
 			       "skipping roll-forward recovery\n");
 			goto skip_recovery;
 		}
+		features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) &
+			~NILFS_FEATURE_COMPAT_RO_SUPP;
+		if (features) {
+			printk(KERN_ERR "NILFS: couldn't proceed with "
+			       "recovery because of unsupported optional "
+			       "features (%llx)\n",
+			       (unsigned long long)features);
+			err = -EROFS;
+			goto failed_unload;
+		}
 		if (really_read_only) {
 			printk(KERN_ERR "NILFS: write access "
 			       "unavailable, cannot proceed.\n");
@@ -320,14 +416,13 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi)
 		goto failed_unload;
 	}
 
-	err = nilfs_recover_logical_segments(nilfs, sbi, &ri);
+	err = nilfs_salvage_orphan_logs(nilfs, sbi, &ri);
 	if (err)
 		goto failed_unload;
 
 	down_write(&nilfs->ns_sem);
-	nilfs->ns_mount_state |= NILFS_VALID_FS;
-	nilfs->ns_sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state);
-	err = nilfs_commit_super(sbi, 1);
+	nilfs->ns_mount_state |= NILFS_VALID_FS; /* set "clean" flag */
+	err = nilfs_cleanup_super(sbi);
 	up_write(&nilfs->ns_sem);
 
 	if (err) {
@@ -343,6 +438,10 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi)
 	sbi->s_super->s_flags = s_flags;
 	return 0;
 
+ scan_error:
+	printk(KERN_ERR "NILFS: error searching super root.\n");
+	goto failed;
+
  failed_unload:
 	nilfs_mdt_destroy(nilfs->ns_cpfile);
 	nilfs_mdt_destroy(nilfs->ns_sufile);
@@ -515,8 +614,8 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs,
 		nilfs_swap_super_block(nilfs);
 	}
 
-	nilfs->ns_sbwtime[0] = le64_to_cpu(sbp[0]->s_wtime);
-	nilfs->ns_sbwtime[1] = valid[!swp] ? le64_to_cpu(sbp[1]->s_wtime) : 0;
+	nilfs->ns_sbwcount = 0;
+	nilfs->ns_sbwtime = le64_to_cpu(sbp[0]->s_wtime);
 	nilfs->ns_prot_seq = le64_to_cpu(sbp[valid[1] & !swp]->s_last_seq);
 	*sbpp = sbp[0];
 	return 0;
@@ -557,6 +656,10 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 		if (err)
 			goto out;
 
+		err = nilfs_check_feature_compatibility(sb, sbp);
+		if (err)
+			goto out;
+
 		blocksize = BLOCK_SIZE << le32_to_cpu(sbp->s_log_block_size);
 		if (sb->s_blocksize != blocksize &&
 		    !sb_set_blocksize(sb, blocksize)) {
@@ -568,7 +671,7 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 		goto out;
 	}
 
-	blocksize = sb_min_blocksize(sb, BLOCK_SIZE);
+	blocksize = sb_min_blocksize(sb, NILFS_MIN_BLOCK_SIZE);
 	if (!blocksize) {
 		printk(KERN_ERR "NILFS: unable to set blocksize\n");
 		err = -EINVAL;
@@ -582,7 +685,18 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 	if (err)
 		goto failed_sbh;
 
+	err = nilfs_check_feature_compatibility(sb, sbp);
+	if (err)
+		goto failed_sbh;
+
 	blocksize = BLOCK_SIZE << le32_to_cpu(sbp->s_log_block_size);
+	if (blocksize < NILFS_MIN_BLOCK_SIZE ||
+	    blocksize > NILFS_MAX_BLOCK_SIZE) {
+		printk(KERN_ERR "NILFS: couldn't mount because of unsupported "
+		       "filesystem blocksize %d\n", blocksize);
+		err = -EINVAL;
+		goto failed_sbh;
+	}
 	if (sb->s_blocksize != blocksize) {
 		int hw_blocksize = bdev_logical_block_size(sb->s_bdev);
 
@@ -604,6 +718,7 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 			   when reloading fails. */
 	}
 	nilfs->ns_blocksize_bits = sb->s_blocksize_bits;
+	nilfs->ns_blocksize = blocksize;
 
 	err = nilfs_store_disk_layout(nilfs, sbp);
 	if (err)
@@ -616,23 +731,9 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 	bdi = nilfs->ns_bdev->bd_inode->i_mapping->backing_dev_info;
 	nilfs->ns_bdi = bdi ? : &default_backing_dev_info;
 
-	/* Finding last segment */
-	nilfs->ns_last_pseg = le64_to_cpu(sbp->s_last_pseg);
-	nilfs->ns_last_cno = le64_to_cpu(sbp->s_last_cno);
-	nilfs->ns_last_seq = le64_to_cpu(sbp->s_last_seq);
-
-	nilfs->ns_seg_seq = nilfs->ns_last_seq;
-	nilfs->ns_segnum =
-		nilfs_get_segnum_of_block(nilfs, nilfs->ns_last_pseg);
-	nilfs->ns_cno = nilfs->ns_last_cno + 1;
-	if (nilfs->ns_segnum >= nilfs->ns_nsegments) {
-		printk(KERN_ERR "NILFS invalid last segment number.\n");
-		err = -EINVAL;
+	err = nilfs_store_log_cursor(nilfs, sbp);
+	if (err)
 		goto failed_sbh;
-	}
-	/* Dummy values  */
-	nilfs->ns_free_segments_count =
-		nilfs->ns_nsegments - (nilfs->ns_segnum + 1);
 
 	/* Initialize gcinode cache */
 	err = nilfs_init_gccache(nilfs);
diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h
index 1ab97453369..f785a7b0ab9 100644
--- a/fs/nilfs2/the_nilfs.h
+++ b/fs/nilfs2/the_nilfs.h
@@ -57,7 +57,8 @@ enum {
  * @ns_current: back pointer to current mount
  * @ns_sbh: buffer heads of on-disk super blocks
  * @ns_sbp: pointers to super block data
- * @ns_sbwtime: previous write time of super blocks
+ * @ns_sbwtime: previous write time of super block
+ * @ns_sbwcount: write count of super block
  * @ns_sbsize: size of valid data in super block
  * @ns_supers: list of nilfs super block structs
  * @ns_seg_seq: segment sequence counter
@@ -73,7 +74,7 @@ enum {
  * @ns_last_seq: sequence value of the latest segment
  * @ns_last_cno: checkpoint number of the latest segment
  * @ns_prot_seq: least sequence number of segments which must not be reclaimed
- * @ns_free_segments_count: counter of free segments
+ * @ns_prev_seq: base sequence number used to decide if advance log cursor
  * @ns_segctor_sem: segment constructor semaphore
  * @ns_dat: DAT file inode
  * @ns_cpfile: checkpoint file inode
@@ -82,6 +83,7 @@ enum {
  * @ns_gc_inodes: dummy inodes to keep live blocks
  * @ns_gc_inodes_h: hash list to keep dummy inode holding live blocks
  * @ns_blocksize_bits: bit length of block size
+ * @ns_blocksize: block size
  * @ns_nsegments: number of segments in filesystem
  * @ns_blocks_per_segment: number of blocks per segment
  * @ns_r_segments_percentage: reserved segments percentage
@@ -119,7 +121,8 @@ struct the_nilfs {
 	 */
 	struct buffer_head     *ns_sbh[2];
 	struct nilfs_super_block *ns_sbp[2];
-	time_t			ns_sbwtime[2];
+	time_t			ns_sbwtime;
+	unsigned		ns_sbwcount;
 	unsigned		ns_sbsize;
 	unsigned		ns_mount_state;
 
@@ -149,7 +152,7 @@ struct the_nilfs {
 	u64			ns_last_seq;
 	__u64			ns_last_cno;
 	u64			ns_prot_seq;
-	unsigned long		ns_free_segments_count;
+	u64			ns_prev_seq;
 
 	struct rw_semaphore	ns_segctor_sem;
 
@@ -168,6 +171,7 @@ struct the_nilfs {
 
 	/* Disk layout information (static) */
 	unsigned int		ns_blocksize_bits;
+	unsigned int		ns_blocksize;
 	unsigned long		ns_nsegments;
 	unsigned long		ns_blocks_per_segment;
 	unsigned long		ns_r_segments_percentage;
@@ -203,20 +207,17 @@ THE_NILFS_FNS(SB_DIRTY, sb_dirty)
 
 /* Minimum interval of periodical update of superblocks (in seconds) */
 #define NILFS_SB_FREQ		10
-#define NILFS_ALTSB_FREQ	60  /* spare superblock */
 
 static inline int nilfs_sb_need_update(struct the_nilfs *nilfs)
 {
 	u64 t = get_seconds();
-	return t < nilfs->ns_sbwtime[0] ||
-		 t > nilfs->ns_sbwtime[0] + NILFS_SB_FREQ;
+	return t < nilfs->ns_sbwtime || t > nilfs->ns_sbwtime + NILFS_SB_FREQ;
 }
 
-static inline int nilfs_altsb_need_update(struct the_nilfs *nilfs)
+static inline int nilfs_sb_will_flip(struct the_nilfs *nilfs)
 {
-	u64 t = get_seconds();
-	struct nilfs_super_block **sbp = nilfs->ns_sbp;
-	return sbp[1] && t > nilfs->ns_sbwtime[1] + NILFS_ALTSB_FREQ;
+	int flip_bits = nilfs->ns_sbwcount & 0x0FL;
+	return (flip_bits != 0x08 && flip_bits != 0x0F);
 }
 
 void nilfs_set_last_segment(struct the_nilfs *, sector_t, u64, __u64);
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 625de9d7088..9b57c0350ff 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -760,13 +760,13 @@ void ocfs2_set_journal_params(struct ocfs2_super *osb)
 	if (osb->osb_commit_interval)
 		commit_interval = osb->osb_commit_interval;
 
-	spin_lock(&journal->j_state_lock);
+	write_lock(&journal->j_state_lock);
 	journal->j_commit_interval = commit_interval;
 	if (osb->s_mount_opt & OCFS2_MOUNT_BARRIER)
 		journal->j_flags |= JBD2_BARRIER;
 	else
 		journal->j_flags &= ~JBD2_BARRIER;
-	spin_unlock(&journal->j_state_lock);
+	write_unlock(&journal->j_state_lock);
 }
 
 int ocfs2_journal_init(struct ocfs2_journal *journal, int *dirty)
diff --git a/fs/partitions/check.c b/fs/partitions/check.c
index 5dcd4b0c553..72c52656dc2 100644
--- a/fs/partitions/check.c
+++ b/fs/partitions/check.c
@@ -459,7 +459,6 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno,
 	}
 
 	/* everything is up and running, commence */
-	INIT_RCU_HEAD(&p->rcu_head);
 	rcu_assign_pointer(ptbl->part[partno], p);
 
 	/* suppress uevent if the disk supresses it */
diff --git a/fs/proc/base.c b/fs/proc/base.c
index acb7ef80ea4..69254a365ce 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -63,6 +63,7 @@
 #include <linux/namei.h>
 #include <linux/mnt_namespace.h>
 #include <linux/mm.h>
+#include <linux/swap.h>
 #include <linux/rcupdate.h>
 #include <linux/kallsyms.h>
 #include <linux/stacktrace.h>
@@ -427,17 +428,14 @@ static const struct file_operations proc_lstats_operations = {
 
 #endif
 
-/* The badness from the OOM killer */
-unsigned long badness(struct task_struct *p, unsigned long uptime);
 static int proc_oom_score(struct task_struct *task, char *buffer)
 {
 	unsigned long points = 0;
-	struct timespec uptime;
 
-	do_posix_clock_monotonic_gettime(&uptime);
 	read_lock(&tasklist_lock);
 	if (pid_alive(task))
-		points = badness(task, uptime.tv_sec);
+		points = oom_badness(task, NULL, NULL,
+					totalram_pages + total_swap_pages);
 	read_unlock(&tasklist_lock);
 	return sprintf(buffer, "%lu\n", points);
 }
@@ -1039,8 +1037,24 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf,
 		return -EACCES;
 	}
 
+	/*
+	 * Warn that /proc/pid/oom_adj is deprecated, see
+	 * Documentation/feature-removal-schedule.txt.
+	 */
+	printk_once(KERN_WARNING "%s (%d): /proc/%d/oom_adj is deprecated, "
+			"please use /proc/%d/oom_score_adj instead.\n",
+			current->comm, task_pid_nr(current),
+			task_pid_nr(task), task_pid_nr(task));
 	task->signal->oom_adj = oom_adjust;
-
+	/*
+	 * Scale /proc/pid/oom_score_adj appropriately ensuring that a maximum
+	 * value is always attainable.
+	 */
+	if (task->signal->oom_adj == OOM_ADJUST_MAX)
+		task->signal->oom_score_adj = OOM_SCORE_ADJ_MAX;
+	else
+		task->signal->oom_score_adj = (oom_adjust * OOM_SCORE_ADJ_MAX) /
+								-OOM_DISABLE;
 	unlock_task_sighand(task, &flags);
 	put_task_struct(task);
 
@@ -1053,6 +1067,82 @@ static const struct file_operations proc_oom_adjust_operations = {
 	.llseek		= generic_file_llseek,
 };
 
+static ssize_t oom_score_adj_read(struct file *file, char __user *buf,
+					size_t count, loff_t *ppos)
+{
+	struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode);
+	char buffer[PROC_NUMBUF];
+	int oom_score_adj = OOM_SCORE_ADJ_MIN;
+	unsigned long flags;
+	size_t len;
+
+	if (!task)
+		return -ESRCH;
+	if (lock_task_sighand(task, &flags)) {
+		oom_score_adj = task->signal->oom_score_adj;
+		unlock_task_sighand(task, &flags);
+	}
+	put_task_struct(task);
+	len = snprintf(buffer, sizeof(buffer), "%d\n", oom_score_adj);
+	return simple_read_from_buffer(buf, count, ppos, buffer, len);
+}
+
+static ssize_t oom_score_adj_write(struct file *file, const char __user *buf,
+					size_t count, loff_t *ppos)
+{
+	struct task_struct *task;
+	char buffer[PROC_NUMBUF];
+	unsigned long flags;
+	long oom_score_adj;
+	int err;
+
+	memset(buffer, 0, sizeof(buffer));
+	if (count > sizeof(buffer) - 1)
+		count = sizeof(buffer) - 1;
+	if (copy_from_user(buffer, buf, count))
+		return -EFAULT;
+
+	err = strict_strtol(strstrip(buffer), 0, &oom_score_adj);
+	if (err)
+		return -EINVAL;
+	if (oom_score_adj < OOM_SCORE_ADJ_MIN ||
+			oom_score_adj > OOM_SCORE_ADJ_MAX)
+		return -EINVAL;
+
+	task = get_proc_task(file->f_path.dentry->d_inode);
+	if (!task)
+		return -ESRCH;
+	if (!lock_task_sighand(task, &flags)) {
+		put_task_struct(task);
+		return -ESRCH;
+	}
+	if (oom_score_adj < task->signal->oom_score_adj &&
+			!capable(CAP_SYS_RESOURCE)) {
+		unlock_task_sighand(task, &flags);
+		put_task_struct(task);
+		return -EACCES;
+	}
+
+	task->signal->oom_score_adj = oom_score_adj;
+	/*
+	 * Scale /proc/pid/oom_adj appropriately ensuring that OOM_DISABLE is
+	 * always attainable.
+	 */
+	if (task->signal->oom_score_adj == OOM_SCORE_ADJ_MIN)
+		task->signal->oom_adj = OOM_DISABLE;
+	else
+		task->signal->oom_adj = (oom_score_adj * OOM_ADJUST_MAX) /
+							OOM_SCORE_ADJ_MAX;
+	unlock_task_sighand(task, &flags);
+	put_task_struct(task);
+	return count;
+}
+
+static const struct file_operations proc_oom_score_adj_operations = {
+	.read		= oom_score_adj_read,
+	.write		= oom_score_adj_write,
+};
+
 #ifdef CONFIG_AUDITSYSCALL
 #define TMPBUFLEN 21
 static ssize_t proc_loginuid_read(struct file * file, char __user * buf,
@@ -2625,6 +2715,7 @@ static const struct pid_entry tgid_base_stuff[] = {
 #endif
 	INF("oom_score",  S_IRUGO, proc_oom_score),
 	REG("oom_adj",    S_IRUGO|S_IWUSR, proc_oom_adjust_operations),
+	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
 #ifdef CONFIG_AUDITSYSCALL
 	REG("loginuid",   S_IWUSR|S_IRUGO, proc_loginuid_operations),
 	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
@@ -2959,6 +3050,7 @@ static const struct pid_entry tid_base_stuff[] = {
 #endif
 	INF("oom_score", S_IRUGO, proc_oom_score),
 	REG("oom_adj",   S_IRUGO|S_IWUSR, proc_oom_adjust_operations),
+	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
 #ifdef CONFIG_AUDITSYSCALL
 	REG("loginuid",  S_IWUSR|S_IRUGO, proc_loginuid_operations),
 	REG("sessionid",  S_IRUSR, proc_sessionid_operations),
diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 437d2ca2de9..ef72b169942 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -132,6 +132,22 @@ static __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_state_lock);
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_data_lock);
 EXPORT_SYMBOL(dq_data_lock);
 
+void __quota_error(struct super_block *sb, const char *func,
+		  const char *fmt, ...)
+{
+	va_list args;
+
+	if (printk_ratelimit()) {
+		va_start(args, fmt);
+		printk(KERN_ERR "Quota error (device %s): %s: ",
+		       sb->s_id, func);
+		vprintk(fmt, args);
+		printk("\n");
+		va_end(args);
+	}
+}
+EXPORT_SYMBOL(__quota_error);
+
 #if defined(CONFIG_QUOTA_DEBUG) || defined(CONFIG_PRINT_QUOTA_WARNING)
 static char *quotatypes[] = INITQFNAMES;
 #endif
@@ -705,11 +721,8 @@ void dqput(struct dquot *dquot)
 		return;
 #ifdef CONFIG_QUOTA_DEBUG
 	if (!atomic_read(&dquot->dq_count)) {
-		printk("VFS: dqput: trying to free free dquot\n");
-		printk("VFS: device %s, dquot of %s %d\n",
-			dquot->dq_sb->s_id,
-			quotatypes[dquot->dq_type],
-			dquot->dq_id);
+		quota_error(dquot->dq_sb, "trying to free free dquot of %s %d",
+			    quotatypes[dquot->dq_type], dquot->dq_id);
 		BUG();
 	}
 #endif
@@ -732,9 +745,9 @@ we_slept:
 		/* Commit dquot before releasing */
 		ret = dquot->dq_sb->dq_op->write_dquot(dquot);
 		if (ret < 0) {
-			printk(KERN_ERR "VFS: cannot write quota structure on "
-				"device %s (error %d). Quota may get out of "
-				"sync!\n", dquot->dq_sb->s_id, ret);
+			quota_error(dquot->dq_sb, "Can't write quota structure"
+				    " (error %d). Quota may get out of sync!",
+				    ret);
 			/*
 			 * We clear dirty bit anyway, so that we avoid
 			 * infinite loop here
@@ -914,9 +927,9 @@ static void add_dquot_ref(struct super_block *sb, int type)
 
 #ifdef CONFIG_QUOTA_DEBUG
 	if (reserved) {
-		printk(KERN_WARNING "VFS (%s): Writes happened before quota"
-			" was turned on thus quota information is probably "
-			"inconsistent. Please run quotacheck(8).\n", sb->s_id);
+		quota_error(sb, "Writes happened before quota was turned on "
+			"thus quota information is probably inconsistent. "
+			"Please run quotacheck(8)");
 	}
 #endif
 }
@@ -947,7 +960,9 @@ static int remove_inode_dquot_ref(struct inode *inode, int type,
 		if (dqput_blocks(dquot)) {
 #ifdef CONFIG_QUOTA_DEBUG
 			if (atomic_read(&dquot->dq_count) != 1)
-				printk(KERN_WARNING "VFS: Adding dquot with dq_count %d to dispose list.\n", atomic_read(&dquot->dq_count));
+				quota_error(inode->i_sb, "Adding dquot with "
+					    "dq_count %d to dispose list",
+					    atomic_read(&dquot->dq_count));
 #endif
 			spin_lock(&dq_list_lock);
 			/* As dquot must have currently users it can't be on
@@ -986,6 +1001,7 @@ static void remove_dquot_ref(struct super_block *sb, int type,
 		struct list_head *tofree_head)
 {
 	struct inode *inode;
+	int reserved = 0;
 
 	spin_lock(&inode_lock);
 	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
@@ -995,10 +1011,20 @@ static void remove_dquot_ref(struct super_block *sb, int type,
 		 *  only quota pointers and these have separate locking
 		 *  (dqptr_sem).
 		 */
-		if (!IS_NOQUOTA(inode))
+		if (!IS_NOQUOTA(inode)) {
+			if (unlikely(inode_get_rsv_space(inode) > 0))
+				reserved = 1;
 			remove_inode_dquot_ref(inode, type, tofree_head);
+		}
 	}
 	spin_unlock(&inode_lock);
+#ifdef CONFIG_QUOTA_DEBUG
+	if (reserved) {
+		printk(KERN_WARNING "VFS (%s): Writes happened after quota"
+			" was disabled thus quota information is probably "
+			"inconsistent. Please run quotacheck(8).\n", sb->s_id);
+	}
+#endif
 }
 
 /* Gather all references from inodes and drop them */
@@ -1304,6 +1330,15 @@ static int info_bdq_free(struct dquot *dquot, qsize_t space)
 	return QUOTA_NL_NOWARN;
 }
 
+static int dquot_active(const struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+
+	if (IS_NOQUOTA(inode))
+		return 0;
+	return sb_any_quota_loaded(sb) & ~sb_any_quota_suspended(sb);
+}
+
 /*
  * Initialize quota pointers in inode
  *
@@ -1323,7 +1358,7 @@ static void __dquot_initialize(struct inode *inode, int type)
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+	if (!dquot_active(inode))
 		return;
 
 	/* First get references to structures we might need. */
@@ -1507,7 +1542,7 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags)
 	 * First test before acquiring mutex - solves deadlocks when we
 	 * re-enter the quota code and are already holding the mutex
 	 */
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
+	if (!dquot_active(inode)) {
 		inode_incr_space(inode, number, reserve);
 		goto out;
 	}
@@ -1559,7 +1594,7 @@ int dquot_alloc_inode(const struct inode *inode)
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+	if (!dquot_active(inode))
 		return 0;
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
 		warntype[cnt] = QUOTA_NL_NOWARN;
@@ -1596,7 +1631,7 @@ int dquot_claim_space_nodirty(struct inode *inode, qsize_t number)
 {
 	int cnt;
 
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
+	if (!dquot_active(inode)) {
 		inode_claim_rsv_space(inode, number);
 		return 0;
 	}
@@ -1629,7 +1664,7 @@ void __dquot_free_space(struct inode *inode, qsize_t number, int flags)
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
+	if (!dquot_active(inode)) {
 		inode_decr_space(inode, number, reserve);
 		return;
 	}
@@ -1667,7 +1702,7 @@ void dquot_free_inode(const struct inode *inode)
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+	if (!dquot_active(inode))
 		return;
 
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
@@ -1790,7 +1825,7 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr)
 	struct super_block *sb = inode->i_sb;
 	int ret;
 
-	if (!sb_any_quota_active(sb) || IS_NOQUOTA(inode))
+	if (!dquot_active(inode))
 		return 0;
 
 	if (iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid)
@@ -1957,7 +1992,7 @@ int dquot_disable(struct super_block *sb, int type, unsigned int flags)
 				truncate_inode_pages(&toputinode[cnt]->i_data,
 						     0);
 				mutex_unlock(&toputinode[cnt]->i_mutex);
-				mark_inode_dirty(toputinode[cnt]);
+				mark_inode_dirty_sync(toputinode[cnt]);
 			}
 			mutex_unlock(&dqopt->dqonoff_mutex);
 		}
@@ -2270,7 +2305,7 @@ static void do_get_dqblk(struct dquot *dquot, struct fs_disk_quota *di)
 	memset(di, 0, sizeof(*di));
 	di->d_version = FS_DQUOT_VERSION;
 	di->d_flags = dquot->dq_type == USRQUOTA ?
-			XFS_USER_QUOTA : XFS_GROUP_QUOTA;
+			FS_USER_QUOTA : FS_GROUP_QUOTA;
 	di->d_id = dquot->dq_id;
 
 	spin_lock(&dq_data_lock);
diff --git a/fs/quota/quota_tree.c b/fs/quota/quota_tree.c
index 24f03407eeb..9e48874eabc 100644
--- a/fs/quota/quota_tree.c
+++ b/fs/quota/quota_tree.c
@@ -65,8 +65,7 @@ static ssize_t write_blk(struct qtree_mem_dqinfo *info, uint blk, char *buf)
 	ret = sb->s_op->quota_write(sb, info->dqi_type, buf,
 	       info->dqi_usable_bs, blk << info->dqi_blocksize_bits);
 	if (ret != info->dqi_usable_bs) {
-		q_warn(KERN_WARNING "VFS: dquota write failed on "
-			"dev %s\n", sb->s_id);
+		quota_error(sb, "dquota write failed");
 		if (ret >= 0)
 			ret = -EIO;
 	}
@@ -160,9 +159,8 @@ static int remove_free_dqentry(struct qtree_mem_dqinfo *info, char *buf,
 	dh->dqdh_next_free = dh->dqdh_prev_free = cpu_to_le32(0);
 	/* No matter whether write succeeds block is out of list */
 	if (write_blk(info, blk, buf) < 0)
-		q_warn(KERN_ERR
-		       "VFS: Can't write block (%u) with free entries.\n",
-		       blk);
+		quota_error(info->dqi_sb, "Can't write block (%u) "
+			    "with free entries", blk);
 	return 0;
 out_buf:
 	kfree(tmpbuf);
@@ -252,9 +250,8 @@ static uint find_free_dqentry(struct qtree_mem_dqinfo *info,
 	if (le16_to_cpu(dh->dqdh_entries) + 1 >= qtree_dqstr_in_blk(info)) {
 		*err = remove_free_dqentry(info, buf, blk);
 		if (*err < 0) {
-			q_warn(KERN_ERR "VFS: find_free_dqentry(): Can't "
-			       "remove block (%u) from entry free list.\n",
-			       blk);
+			quota_error(dquot->dq_sb, "Can't remove block (%u) "
+				    "from entry free list", blk);
 			goto out_buf;
 		}
 	}
@@ -268,16 +265,15 @@ static uint find_free_dqentry(struct qtree_mem_dqinfo *info,
 	}
 #ifdef __QUOTA_QT_PARANOIA
 	if (i == qtree_dqstr_in_blk(info)) {
-		printk(KERN_ERR "VFS: find_free_dqentry(): Data block full "
-				"but it shouldn't.\n");
+		quota_error(dquot->dq_sb, "Data block full but it shouldn't");
 		*err = -EIO;
 		goto out_buf;
 	}
 #endif
 	*err = write_blk(info, blk, buf);
 	if (*err < 0) {
-		q_warn(KERN_ERR "VFS: find_free_dqentry(): Can't write quota "
-				"data block %u.\n", blk);
+		quota_error(dquot->dq_sb, "Can't write quota data block %u",
+			    blk);
 		goto out_buf;
 	}
 	dquot->dq_off = (blk << info->dqi_blocksize_bits) +
@@ -311,8 +307,8 @@ static int do_insert_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 	} else {
 		ret = read_blk(info, *treeblk, buf);
 		if (ret < 0) {
-			q_warn(KERN_ERR "VFS: Can't read tree quota block "
-					"%u.\n", *treeblk);
+			quota_error(dquot->dq_sb, "Can't read tree quota "
+				    "block %u", *treeblk);
 			goto out_buf;
 		}
 	}
@@ -323,9 +319,9 @@ static int do_insert_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 	if (depth == info->dqi_qtree_depth - 1) {
 #ifdef __QUOTA_QT_PARANOIA
 		if (newblk) {
-			printk(KERN_ERR "VFS: Inserting already present quota "
-					"entry (block %u).\n",
-			       le32_to_cpu(ref[get_index(info,
+			quota_error(dquot->dq_sb, "Inserting already present "
+				    "quota entry (block %u)",
+				    le32_to_cpu(ref[get_index(info,
 						dquot->dq_id, depth)]));
 			ret = -EIO;
 			goto out_buf;
@@ -373,8 +369,8 @@ int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
 	if (!dquot->dq_off) {
 		ret = dq_insert_tree(info, dquot);
 		if (ret < 0) {
-			q_warn(KERN_ERR "VFS: Error %zd occurred while "
-					"creating quota.\n", ret);
+			quota_error(sb, "Error %zd occurred while creating "
+				    "quota", ret);
 			kfree(ddquot);
 			return ret;
 		}
@@ -385,8 +381,7 @@ int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
 	ret = sb->s_op->quota_write(sb, type, ddquot, info->dqi_entry_size,
 				    dquot->dq_off);
 	if (ret != info->dqi_entry_size) {
-		q_warn(KERN_WARNING "VFS: dquota write failed on dev %s\n",
-		       sb->s_id);
+		quota_error(sb, "dquota write failed");
 		if (ret >= 0)
 			ret = -ENOSPC;
 	} else {
@@ -410,14 +405,15 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 	if (!buf)
 		return -ENOMEM;
 	if (dquot->dq_off >> info->dqi_blocksize_bits != blk) {
-		q_warn(KERN_ERR "VFS: Quota structure has offset to other "
-		  "block (%u) than it should (%u).\n", blk,
-		  (uint)(dquot->dq_off >> info->dqi_blocksize_bits));
+		quota_error(dquot->dq_sb, "Quota structure has offset to "
+			"other block (%u) than it should (%u)", blk,
+			(uint)(dquot->dq_off >> info->dqi_blocksize_bits));
 		goto out_buf;
 	}
 	ret = read_blk(info, blk, buf);
 	if (ret < 0) {
-		q_warn(KERN_ERR "VFS: Can't read quota data block %u\n", blk);
+		quota_error(dquot->dq_sb, "Can't read quota data block %u",
+			    blk);
 		goto out_buf;
 	}
 	dh = (struct qt_disk_dqdbheader *)buf;
@@ -427,8 +423,8 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 		if (ret >= 0)
 			ret = put_free_dqblk(info, buf, blk);
 		if (ret < 0) {
-			q_warn(KERN_ERR "VFS: Can't move quota data block (%u) "
-			  "to free list.\n", blk);
+			quota_error(dquot->dq_sb, "Can't move quota data block "
+				    "(%u) to free list", blk);
 			goto out_buf;
 		}
 	} else {
@@ -440,15 +436,15 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 			/* Insert will write block itself */
 			ret = insert_free_dqentry(info, buf, blk);
 			if (ret < 0) {
-				q_warn(KERN_ERR "VFS: Can't insert quota data "
-				       "block (%u) to free entry list.\n", blk);
+				quota_error(dquot->dq_sb, "Can't insert quota "
+				    "data block (%u) to free entry list", blk);
 				goto out_buf;
 			}
 		} else {
 			ret = write_blk(info, blk, buf);
 			if (ret < 0) {
-				q_warn(KERN_ERR "VFS: Can't write quota data "
-				  "block %u\n", blk);
+				quota_error(dquot->dq_sb, "Can't write quota "
+					    "data block %u", blk);
 				goto out_buf;
 			}
 		}
@@ -472,7 +468,8 @@ static int remove_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 		return -ENOMEM;
 	ret = read_blk(info, *blk, buf);
 	if (ret < 0) {
-		q_warn(KERN_ERR "VFS: Can't read quota data block %u\n", *blk);
+		quota_error(dquot->dq_sb, "Can't read quota data "
+			    "block %u", blk);
 		goto out_buf;
 	}
 	newblk = le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)]);
@@ -496,8 +493,8 @@ static int remove_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
 		} else {
 			ret = write_blk(info, *blk, buf);
 			if (ret < 0)
-				q_warn(KERN_ERR "VFS: Can't write quota tree "
-				  "block %u.\n", *blk);
+				quota_error(dquot->dq_sb, "Can't write quota "
+					    "tree block %u", blk);
 		}
 	}
 out_buf:
@@ -529,7 +526,8 @@ static loff_t find_block_dqentry(struct qtree_mem_dqinfo *info,
 		return -ENOMEM;
 	ret = read_blk(info, blk, buf);
 	if (ret < 0) {
-		q_warn(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
+		quota_error(dquot->dq_sb, "Can't read quota tree "
+			    "block %u", blk);
 		goto out_buf;
 	}
 	ddquot = buf + sizeof(struct qt_disk_dqdbheader);
@@ -539,8 +537,8 @@ static loff_t find_block_dqentry(struct qtree_mem_dqinfo *info,
 		ddquot += info->dqi_entry_size;
 	}
 	if (i == qtree_dqstr_in_blk(info)) {
-		q_warn(KERN_ERR "VFS: Quota for id %u referenced "
-		  "but not present.\n", dquot->dq_id);
+		quota_error(dquot->dq_sb, "Quota for id %u referenced "
+			    "but not present", dquot->dq_id);
 		ret = -EIO;
 		goto out_buf;
 	} else {
@@ -564,7 +562,8 @@ static loff_t find_tree_dqentry(struct qtree_mem_dqinfo *info,
 		return -ENOMEM;
 	ret = read_blk(info, blk, buf);
 	if (ret < 0) {
-		q_warn(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
+		quota_error(dquot->dq_sb, "Can't read quota tree block %u",
+			    blk);
 		goto out_buf;
 	}
 	ret = 0;
@@ -598,7 +597,7 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
 #ifdef __QUOTA_QT_PARANOIA
 	/* Invalidated quota? */
 	if (!sb_dqopt(dquot->dq_sb)->files[type]) {
-		printk(KERN_ERR "VFS: Quota invalidated while reading!\n");
+		quota_error(sb, "Quota invalidated while reading!");
 		return -EIO;
 	}
 #endif
@@ -607,8 +606,8 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
 		offset = find_dqentry(info, dquot);
 		if (offset <= 0) {	/* Entry not present? */
 			if (offset < 0)
-				q_warn(KERN_ERR "VFS: Can't read quota "
-				  "structure for id %u.\n", dquot->dq_id);
+				quota_error(sb, "Can't read quota structure "
+					    "for id %u", dquot->dq_id);
 			dquot->dq_off = 0;
 			set_bit(DQ_FAKE_B, &dquot->dq_flags);
 			memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk));
@@ -625,8 +624,8 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
 	if (ret != info->dqi_entry_size) {
 		if (ret >= 0)
 			ret = -EIO;
-		q_warn(KERN_ERR "VFS: Error while reading quota "
-				"structure for id %u.\n", dquot->dq_id);
+		quota_error(sb, "Error while reading quota structure for id %u",
+			    dquot->dq_id);
 		set_bit(DQ_FAKE_B, &dquot->dq_flags);
 		memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk));
 		kfree(ddquot);
diff --git a/fs/quota/quota_tree.h b/fs/quota/quota_tree.h
index ccc3e71fb1d..a1ab8db81a5 100644
--- a/fs/quota/quota_tree.h
+++ b/fs/quota/quota_tree.h
@@ -22,10 +22,4 @@ struct qt_disk_dqdbheader {
 
 #define QT_TREEOFF	1		/* Offset of tree in file in blocks */
 
-#define q_warn(fmt, args...) \
-do { \
-	if (printk_ratelimit()) \
-		printk(fmt, ## args); \
-} while(0)
-
 #endif /* _LINUX_QUOTAIO_TREE_H */
diff --git a/fs/quota/quota_v1.c b/fs/quota/quota_v1.c
index 4af344c5852..34b37a67bb1 100644
--- a/fs/quota/quota_v1.c
+++ b/fs/quota/quota_v1.c
@@ -95,8 +95,7 @@ static int v1_commit_dqblk(struct dquot *dquot)
 			(char *)&dqblk, sizeof(struct v1_disk_dqblk),
 			v1_dqoff(dquot->dq_id));
 	if (ret != sizeof(struct v1_disk_dqblk)) {
-		printk(KERN_WARNING "VFS: dquota write failed on dev %s\n",
-			dquot->dq_sb->s_id);
+		quota_error(dquot->dq_sb, "dquota write failed");
 		if (ret >= 0)
 			ret = -EIO;
 		goto out;
diff --git a/fs/quota/quota_v2.c b/fs/quota/quota_v2.c
index 135206af145..65444d29406 100644
--- a/fs/quota/quota_v2.c
+++ b/fs/quota/quota_v2.c
@@ -63,9 +63,8 @@ static int v2_read_header(struct super_block *sb, int type,
 	size = sb->s_op->quota_read(sb, type, (char *)dqhead,
 				    sizeof(struct v2_disk_dqheader), 0);
 	if (size != sizeof(struct v2_disk_dqheader)) {
-		q_warn(KERN_WARNING "quota_v2: Failed header read:"
-		       " expected=%zd got=%zd\n",
-			sizeof(struct v2_disk_dqheader), size);
+		quota_error(sb, "Failed header read: expected=%zd got=%zd",
+			    sizeof(struct v2_disk_dqheader), size);
 		return 0;
 	}
 	return 1;
@@ -106,8 +105,7 @@ static int v2_read_file_info(struct super_block *sb, int type)
 	size = sb->s_op->quota_read(sb, type, (char *)&dinfo,
 	       sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF);
 	if (size != sizeof(struct v2_disk_dqinfo)) {
-		q_warn(KERN_WARNING "quota_v2: Can't read info structure on device %s.\n",
-			sb->s_id);
+		quota_error(sb, "Can't read info structure");
 		return -1;
 	}
 	info->dqi_priv = kmalloc(sizeof(struct qtree_mem_dqinfo), GFP_NOFS);
@@ -167,8 +165,7 @@ static int v2_write_file_info(struct super_block *sb, int type)
 	size = sb->s_op->quota_write(sb, type, (char *)&dinfo,
 	       sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF);
 	if (size != sizeof(struct v2_disk_dqinfo)) {
-		q_warn(KERN_WARNING "Can't write info structure on device %s.\n",
-			sb->s_id);
+		quota_error(sb, "Can't write info structure");
 		return -1;
 	}
 	return 0;
diff --git a/fs/readdir.c b/fs/readdir.c
index 7723401f8d8..356f71528ad 100644
--- a/fs/readdir.c
+++ b/fs/readdir.c
@@ -4,6 +4,7 @@
  *  Copyright (C) 1995  Linus Torvalds
  */
 
+#include <linux/stddef.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/time.h>
@@ -54,7 +55,6 @@ EXPORT_SYMBOL(vfs_readdir);
  * anyway. Thus the special "fillonedir()" function for that
  * case (the low-level handlers don't need to care about this).
  */
-#define NAME_OFFSET(de) ((int) ((de)->d_name - (char __user *) (de)))
 
 #ifdef __ARCH_WANT_OLD_READDIR
 
@@ -152,7 +152,8 @@ static int filldir(void * __buf, const char * name, int namlen, loff_t offset,
 	struct linux_dirent __user * dirent;
 	struct getdents_callback * buf = (struct getdents_callback *) __buf;
 	unsigned long d_ino;
-	int reclen = ALIGN(NAME_OFFSET(dirent) + namlen + 2, sizeof(long));
+	int reclen = ALIGN(offsetof(struct linux_dirent, d_name) + namlen + 2,
+		sizeof(long));
 
 	buf->error = -EINVAL;	/* only used if we fail.. */
 	if (reclen > buf->count)
@@ -237,7 +238,8 @@ static int filldir64(void * __buf, const char * name, int namlen, loff_t offset,
 {
 	struct linux_dirent64 __user *dirent;
 	struct getdents_callback64 * buf = (struct getdents_callback64 *) __buf;
-	int reclen = ALIGN(NAME_OFFSET(dirent) + namlen + 1, sizeof(u64));
+	int reclen = ALIGN(offsetof(struct linux_dirent64, d_name) + namlen + 1,
+		sizeof(u64));
 
 	buf->error = -EINVAL;	/* only used if we fail.. */
 	if (reclen > buf->count)
diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c
index 1beaa739d0a..1b27b5688f6 100644
--- a/fs/sysfs/file.c
+++ b/fs/sysfs/file.c
@@ -593,7 +593,8 @@ EXPORT_SYMBOL_GPL(sysfs_add_file_to_group);
  * @mode: file permissions.
  *
  */
-int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode)
+int sysfs_chmod_file(struct kobject *kobj, const struct attribute *attr,
+		     mode_t mode)
 {
 	struct sysfs_dirent *sd;
 	struct iattr newattrs;
diff --git a/fs/udf/file.c b/fs/udf/file.c
index 94e06d6bddb..6e450e01a1b 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -36,7 +36,6 @@
 #include <linux/pagemap.h>
 #include <linux/buffer_head.h>
 #include <linux/aio.h>
-#include <linux/smp_lock.h>
 
 #include "udf_i.h"
 #include "udf_sb.h"
diff --git a/fs/udf/super.c b/fs/udf/super.c
index 612d1e2e285..12bb651e540 100644
--- a/fs/udf/super.c
+++ b/fs/udf/super.c
@@ -1579,9 +1579,7 @@ static int udf_load_sequence(struct super_block *sb, struct buffer_head *bh,
 {
 	struct anchorVolDescPtr *anchor;
 	long main_s, main_e, reserve_s, reserve_e;
-	struct udf_sb_info *sbi;
 
-	sbi = UDF_SB(sb);
 	anchor = (struct anchorVolDescPtr *)bh->b_data;
 
 	/* Locate the main sequence */
diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c
index bfd5ac9d1f6..29b9d642e93 100644
--- a/fs/xfs/linux-2.6/xfs_quotaops.c
+++ b/fs/xfs/linux-2.6/xfs_quotaops.c
@@ -68,15 +68,15 @@ xfs_fs_set_xstate(
 	if (op != Q_XQUOTARM && !XFS_IS_QUOTA_RUNNING(mp))
 		return -ENOSYS;
 
-	if (uflags & XFS_QUOTA_UDQ_ACCT)
+	if (uflags & FS_QUOTA_UDQ_ACCT)
 		flags |= XFS_UQUOTA_ACCT;
-	if (uflags & XFS_QUOTA_PDQ_ACCT)
+	if (uflags & FS_QUOTA_PDQ_ACCT)
 		flags |= XFS_PQUOTA_ACCT;
-	if (uflags & XFS_QUOTA_GDQ_ACCT)
+	if (uflags & FS_QUOTA_GDQ_ACCT)
 		flags |= XFS_GQUOTA_ACCT;
-	if (uflags & XFS_QUOTA_UDQ_ENFD)
+	if (uflags & FS_QUOTA_UDQ_ENFD)
 		flags |= XFS_UQUOTA_ENFD;
-	if (uflags & (XFS_QUOTA_PDQ_ENFD|XFS_QUOTA_GDQ_ENFD))
+	if (uflags & (FS_QUOTA_PDQ_ENFD|FS_QUOTA_GDQ_ENFD))
 		flags |= XFS_OQUOTA_ENFD;
 
 	switch (op) {
diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c
index d257eb8557c..45e5849df23 100644
--- a/fs/xfs/quota/xfs_qm_syscalls.c
+++ b/fs/xfs/quota/xfs_qm_syscalls.c
@@ -810,9 +810,9 @@ xfs_qm_export_dquot(
 	}
 
 #ifdef DEBUG
-	if (((XFS_IS_UQUOTA_ENFORCED(mp) && dst->d_flags == XFS_USER_QUOTA) ||
+	if (((XFS_IS_UQUOTA_ENFORCED(mp) && dst->d_flags == FS_USER_QUOTA) ||
 	     (XFS_IS_OQUOTA_ENFORCED(mp) &&
-			(dst->d_flags & (XFS_PROJ_QUOTA | XFS_GROUP_QUOTA)))) &&
+			(dst->d_flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)))) &&
 	    dst->d_id != 0) {
 		if (((int) dst->d_bcount >= (int) dst->d_blk_softlimit) &&
 		    (dst->d_blk_softlimit > 0)) {
@@ -833,17 +833,17 @@ xfs_qm_export_qtype_flags(
 	/*
 	 * Can't be more than one, or none.
 	 */
-	ASSERT((flags & (XFS_PROJ_QUOTA | XFS_USER_QUOTA)) !=
-		(XFS_PROJ_QUOTA | XFS_USER_QUOTA));
-	ASSERT((flags & (XFS_PROJ_QUOTA | XFS_GROUP_QUOTA)) !=
-		(XFS_PROJ_QUOTA | XFS_GROUP_QUOTA));
-	ASSERT((flags & (XFS_USER_QUOTA | XFS_GROUP_QUOTA)) !=
-		(XFS_USER_QUOTA | XFS_GROUP_QUOTA));
-	ASSERT((flags & (XFS_PROJ_QUOTA|XFS_USER_QUOTA|XFS_GROUP_QUOTA)) != 0);
+	ASSERT((flags & (FS_PROJ_QUOTA | FS_USER_QUOTA)) !=
+		(FS_PROJ_QUOTA | FS_USER_QUOTA));
+	ASSERT((flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)) !=
+		(FS_PROJ_QUOTA | FS_GROUP_QUOTA));
+	ASSERT((flags & (FS_USER_QUOTA | FS_GROUP_QUOTA)) !=
+		(FS_USER_QUOTA | FS_GROUP_QUOTA));
+	ASSERT((flags & (FS_PROJ_QUOTA|FS_USER_QUOTA|FS_GROUP_QUOTA)) != 0);
 
 	return (flags & XFS_DQ_USER) ?
-		XFS_USER_QUOTA : (flags & XFS_DQ_PROJ) ?
-			XFS_PROJ_QUOTA : XFS_GROUP_QUOTA;
+		FS_USER_QUOTA : (flags & XFS_DQ_PROJ) ?
+			FS_PROJ_QUOTA : FS_GROUP_QUOTA;
 }
 
 STATIC uint
@@ -854,16 +854,16 @@ xfs_qm_export_flags(
 
 	uflags = 0;
 	if (flags & XFS_UQUOTA_ACCT)
-		uflags |= XFS_QUOTA_UDQ_ACCT;
+		uflags |= FS_QUOTA_UDQ_ACCT;
 	if (flags & XFS_PQUOTA_ACCT)
-		uflags |= XFS_QUOTA_PDQ_ACCT;
+		uflags |= FS_QUOTA_PDQ_ACCT;
 	if (flags & XFS_GQUOTA_ACCT)
-		uflags |= XFS_QUOTA_GDQ_ACCT;
+		uflags |= FS_QUOTA_GDQ_ACCT;
 	if (flags & XFS_UQUOTA_ENFD)
-		uflags |= XFS_QUOTA_UDQ_ENFD;
+		uflags |= FS_QUOTA_UDQ_ENFD;
 	if (flags & (XFS_OQUOTA_ENFD)) {
 		uflags |= (flags & XFS_GQUOTA_ACCT) ?
-			XFS_QUOTA_GDQ_ENFD : XFS_QUOTA_PDQ_ENFD;
+			FS_QUOTA_GDQ_ENFD : FS_QUOTA_PDQ_ENFD;
 	}
 	return (uflags);
 }