Merge HEAD from ../linux-2.6

40 files changed, 585 insertions, 272 deletions

diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c
index ae6d032b9b59..cd5eeb032d64 100644
--- a/fs/9p/vfs_dir.c
+++ b/fs/9p/vfs_dir.c
@@ -202,7 +202,6 @@ int v9fs_dir_release(struct inode *inode, struct file *filp)
 		filp->private_data = NULL;
 	}
 
-	d_drop(filp->f_dentry);
 	return 0;
 }
 
diff --git a/fs/buffer.c b/fs/buffer.c
index 62cfd17dc5fe..a9b399402007 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -3060,6 +3060,7 @@ int buffer_migrate_page(struct page *newpage, struct page *page)
 {
 	struct address_space *mapping = page->mapping;
 	struct buffer_head *bh, *head;
+	int rc;
 
 	if (!mapping)
 		return -EAGAIN;
@@ -3069,8 +3070,9 @@ int buffer_migrate_page(struct page *newpage, struct page *page)
 
 	head = page_buffers(page);
 
-	if (migrate_page_remove_references(newpage, page, 3))
-		return -EAGAIN;
+	rc = migrate_page_remove_references(newpage, page, 3);
+	if (rc)
+		return rc;
 
 	bh = head;
 	do {
diff --git a/fs/char_dev.c b/fs/char_dev.c
index 21195c481637..5c36345c9bf7 100644
--- a/fs/char_dev.c
+++ b/fs/char_dev.c
@@ -19,6 +19,7 @@
 #include <linux/kobject.h>
 #include <linux/kobj_map.h>
 #include <linux/cdev.h>
+#include <linux/mutex.h>
 
 #ifdef CONFIG_KMOD
 #include <linux/kmod.h>
@@ -28,7 +29,7 @@ static struct kobj_map *cdev_map;
 
 #define MAX_PROBE_HASH 255	/* random */
 
-static DECLARE_MUTEX(chrdevs_lock);
+static DEFINE_MUTEX(chrdevs_lock);
 
 static struct char_device_struct {
 	struct char_device_struct *next;
@@ -88,13 +89,13 @@ out:
 
 void *acquire_chrdev_list(void)
 {
-	down(&chrdevs_lock);
+	mutex_lock(&chrdevs_lock);
 	return get_next_chrdev(NULL);
 }
 
 void release_chrdev_list(void *dev)
 {
-	up(&chrdevs_lock);
+	mutex_unlock(&chrdevs_lock);
 	kfree(dev);
 }
 
@@ -151,7 +152,7 @@ __register_chrdev_region(unsigned int major, unsigned int baseminor,
 
 	memset(cd, 0, sizeof(struct char_device_struct));
 
-	down(&chrdevs_lock);
+	mutex_lock(&chrdevs_lock);
 
 	/* temporary */
 	if (major == 0) {
@@ -186,10 +187,10 @@ __register_chrdev_region(unsigned int major, unsigned int baseminor,
 	}
 	cd->next = *cp;
 	*cp = cd;
-	up(&chrdevs_lock);
+	mutex_unlock(&chrdevs_lock);
 	return cd;
 out:
-	up(&chrdevs_lock);
+	mutex_unlock(&chrdevs_lock);
 	kfree(cd);
 	return ERR_PTR(ret);
 }
@@ -200,7 +201,7 @@ __unregister_chrdev_region(unsigned major, unsigned baseminor, int minorct)
 	struct char_device_struct *cd = NULL, **cp;
 	int i = major_to_index(major);
 
-	down(&chrdevs_lock);
+	mutex_lock(&chrdevs_lock);
 	for (cp = &chrdevs[i]; *cp; cp = &(*cp)->next)
 		if ((*cp)->major == major &&
 		    (*cp)->baseminor == baseminor &&
@@ -210,7 +211,7 @@ __unregister_chrdev_region(unsigned major, unsigned baseminor, int minorct)
 		cd = *cp;
 		*cp = cd->next;
 	}
-	up(&chrdevs_lock);
+	mutex_unlock(&chrdevs_lock);
 	return cd;
 }
 
diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index d575452cd9f7..40c4fc973fad 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -251,3 +251,49 @@ struct dentry *debugfs_create_bool(const char *name, mode_t mode,
 }
 EXPORT_SYMBOL_GPL(debugfs_create_bool);
 
+static ssize_t read_file_blob(struct file *file, char __user *user_buf,
+			      size_t count, loff_t *ppos)
+{
+	struct debugfs_blob_wrapper *blob = file->private_data;
+	return simple_read_from_buffer(user_buf, count, ppos, blob->data,
+			blob->size);
+}
+
+static struct file_operations fops_blob = {
+	.read =		read_file_blob,
+	.open =		default_open,
+};
+
+/**
+ * debugfs_create_blob - create a file in the debugfs filesystem that is
+ * used to read and write a binary blob.
+ *
+ * @name: a pointer to a string containing the name of the file to create.
+ * @mode: the permission that the file should have
+ * @parent: a pointer to the parent dentry for this file.  This should be a
+ *          directory dentry if set.  If this paramater is NULL, then the
+ *          file will be created in the root of the debugfs filesystem.
+ * @blob: a pointer to a struct debugfs_blob_wrapper which contains a pointer
+ *        to the blob data and the size of the data.
+ *
+ * This function creates a file in debugfs with the given name that exports
+ * @blob->data as a binary blob. If the @mode variable is so set it can be
+ * read from. Writing is not supported.
+ *
+ * This function will return a pointer to a dentry if it succeeds.  This
+ * pointer must be passed to the debugfs_remove() function when the file is
+ * to be removed (no automatic cleanup happens if your module is unloaded,
+ * you are responsible here.)  If an error occurs, NULL will be returned.
+ *
+ * If debugfs is not enabled in the kernel, the value -ENODEV will be
+ * returned.  It is not wise to check for this value, but rather, check for
+ * NULL or !NULL instead as to eliminate the need for #ifdef in the calling
+ * code.
+ */
+struct dentry *debugfs_create_blob(const char *name, mode_t mode,
+				   struct dentry *parent,
+				   struct debugfs_blob_wrapper *blob)
+{
+	return debugfs_create_file(name, mode, parent, blob, &fops_blob);
+}
+EXPORT_SYMBOL_GPL(debugfs_create_blob);
diff --git a/fs/direct-io.c b/fs/direct-io.c
index 848044af7e16..27f3e787faca 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -1155,15 +1155,16 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode,
  * For writes, i_mutex is not held on entry; it is never taken.
  *
  * DIO_LOCKING (simple locking for regular files)
- * For writes we are called under i_mutex and return with i_mutex held, even though
- * it is internally dropped.
+ * For writes we are called under i_mutex and return with i_mutex held, even
+ * though it is internally dropped.
  * For reads, i_mutex is not held on entry, but it is taken and dropped before
  * returning.
  *
  * DIO_OWN_LOCKING (filesystem provides synchronisation and handling of
  *	uninitialised data, allowing parallel direct readers and writers)
  * For writes we are called without i_mutex, return without it, never touch it.
- * For reads, i_mutex is held on entry and will be released before returning.
+ * For reads we are called under i_mutex and return with i_mutex held, even
+ * though it may be internally dropped.
  *
  * Additional i_alloc_sem locking requirements described inline below.
  */
@@ -1182,7 +1183,8 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 	ssize_t retval = -EINVAL;
 	loff_t end = offset;
 	struct dio *dio;
-	int reader_with_isem = (rw == READ && dio_lock_type == DIO_OWN_LOCKING);
+	int release_i_mutex = 0;
+	int acquire_i_mutex = 0;
 
 	if (rw & WRITE)
 		current->flags |= PF_SYNCWRITE;
@@ -1225,7 +1227,6 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 	 *	writers need to grab i_alloc_sem only (i_mutex is already held)
 	 * For regular files using DIO_OWN_LOCKING,
 	 *	neither readers nor writers take any locks here
-	 *	(i_mutex is already held and release for writers here)
 	 */
 	dio->lock_type = dio_lock_type;
 	if (dio_lock_type != DIO_NO_LOCKING) {
@@ -1236,7 +1237,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 			mapping = iocb->ki_filp->f_mapping;
 			if (dio_lock_type != DIO_OWN_LOCKING) {
 				mutex_lock(&inode->i_mutex);
-				reader_with_isem = 1;
+				release_i_mutex = 1;
 			}
 
 			retval = filemap_write_and_wait_range(mapping, offset,
@@ -1248,7 +1249,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 
 			if (dio_lock_type == DIO_OWN_LOCKING) {
 				mutex_unlock(&inode->i_mutex);
-				reader_with_isem = 0;
+				acquire_i_mutex = 1;
 			}
 		}
 
@@ -1269,11 +1270,13 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 				nr_segs, blkbits, get_blocks, end_io, dio);
 
 	if (rw == READ && dio_lock_type == DIO_LOCKING)
-		reader_with_isem = 0;
+		release_i_mutex = 0;
 
 out:
-	if (reader_with_isem)
+	if (release_i_mutex)
 		mutex_unlock(&inode->i_mutex);
+	else if (acquire_i_mutex)
+		mutex_lock(&inode->i_mutex);
 	if (rw & WRITE)
 		current->flags &= ~PF_SYNCWRITE;
 	return retval;
diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c
index 7442bdd1267a..b3dbd716cd3a 100644
--- a/fs/ext2/dir.c
+++ b/fs/ext2/dir.c
@@ -256,11 +256,10 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir)
 	unsigned long npages = dir_pages(inode);
 	unsigned chunk_mask = ~(ext2_chunk_size(inode)-1);
 	unsigned char *types = NULL;
-	int need_revalidate = (filp->f_version != inode->i_version);
-	int ret;
+	int need_revalidate = filp->f_version != inode->i_version;
 
 	if (pos > inode->i_size - EXT2_DIR_REC_LEN(1))
-		goto success;
+		return 0;
 
 	if (EXT2_HAS_INCOMPAT_FEATURE(sb, EXT2_FEATURE_INCOMPAT_FILETYPE))
 		types = ext2_filetype_table;
@@ -275,12 +274,15 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir)
 				   "bad page in #%lu",
 				   inode->i_ino);
 			filp->f_pos += PAGE_CACHE_SIZE - offset;
-			ret = -EIO;
-			goto done;
+			return -EIO;
 		}
 		kaddr = page_address(page);
-		if (need_revalidate) {
-			offset = ext2_validate_entry(kaddr, offset, chunk_mask);
+		if (unlikely(need_revalidate)) {
+			if (offset) {
+				offset = ext2_validate_entry(kaddr, offset, chunk_mask);
+				filp->f_pos = (n<<PAGE_CACHE_SHIFT) + offset;
+			}
+			filp->f_version = inode->i_version;
 			need_revalidate = 0;
 		}
 		de = (ext2_dirent *)(kaddr+offset);
@@ -289,9 +291,8 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir)
 			if (de->rec_len == 0) {
 				ext2_error(sb, __FUNCTION__,
 					"zero-length directory entry");
-				ret = -EIO;
 				ext2_put_page(page);
-				goto done;
+				return -EIO;
 			}
 			if (de->inode) {
 				int over;
@@ -306,19 +307,14 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir)
 						le32_to_cpu(de->inode), d_type);
 				if (over) {
 					ext2_put_page(page);
-					goto success;
+					return 0;
 				}
 			}
 			filp->f_pos += le16_to_cpu(de->rec_len);
 		}
 		ext2_put_page(page);
 	}
-
-success:
-	ret = 0;
-done:
-	filp->f_version = inode->i_version;
-	return ret;
+	return 0;
 }
 
 /*
diff --git a/fs/jfs/Makefile b/fs/jfs/Makefile
index 6f1e0e95587a..3adb6395e42d 100644
--- a/fs/jfs/Makefile
+++ b/fs/jfs/Makefile
@@ -8,7 +8,8 @@ jfs-y    := super.o file.o inode.o namei.o jfs_mount.o jfs_umount.o \
 	    jfs_xtree.o jfs_imap.o jfs_debug.o jfs_dmap.o \
 	    jfs_unicode.o jfs_dtree.o jfs_inode.o \
 	    jfs_extent.o symlink.o jfs_metapage.o \
-	    jfs_logmgr.o jfs_txnmgr.o jfs_uniupr.o resize.o xattr.o
+	    jfs_logmgr.o jfs_txnmgr.o jfs_uniupr.o \
+	    resize.o xattr.o ioctl.o
 
 jfs-$(CONFIG_JFS_POSIX_ACL) += acl.o
 
diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c
index 461e4934ca7c..e2281300979c 100644
--- a/fs/jfs/acl.c
+++ b/fs/jfs/acl.c
@@ -183,6 +183,9 @@ cleanup:
 		posix_acl_release(acl);
 	} else
 		inode->i_mode &= ~current->fs->umask;
+	
+	JFS_IP(inode)->mode2 = (JFS_IP(inode)->mode2 & 0xffff0000) |
+			       inode->i_mode;
 
 	return rc;
 }
@@ -207,12 +210,12 @@ static int jfs_acl_chmod(struct inode *inode)
 	rc = posix_acl_chmod_masq(clone, inode->i_mode);
 	if (!rc) {
 		tid_t tid = txBegin(inode->i_sb, 0);
-		down(&JFS_IP(inode)->commit_sem);
+		mutex_lock(&JFS_IP(inode)->commit_mutex);
 		rc = jfs_set_acl(tid, inode, ACL_TYPE_ACCESS, clone);
 		if (!rc)
 			rc = txCommit(tid, 1, &inode, 0);
 		txEnd(tid);
-		up(&JFS_IP(inode)->commit_sem);
+		mutex_unlock(&JFS_IP(inode)->commit_mutex);
 	}
 
 	posix_acl_release(clone);
diff --git a/fs/jfs/file.c b/fs/jfs/file.c
index c2c19c9ed9a4..e1ac6e497e2b 100644
--- a/fs/jfs/file.c
+++ b/fs/jfs/file.c
@@ -113,4 +113,5 @@ struct file_operations jfs_file_operations = {
  	.sendfile	= generic_file_sendfile,
 	.fsync		= jfs_fsync,
 	.release	= jfs_release,
+	.ioctl		= jfs_ioctl,
 };
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index 9f942ca8e4e3..51a5fed90cca 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -55,6 +55,7 @@ void jfs_read_inode(struct inode *inode)
 		inode->i_op = &jfs_file_inode_operations;
 		init_special_inode(inode, inode->i_mode, inode->i_rdev);
 	}
+	jfs_set_inode_flags(inode);
 }
 
 /*
@@ -89,16 +90,16 @@ int jfs_commit_inode(struct inode *inode, int wait)
 	}
 
 	tid = txBegin(inode->i_sb, COMMIT_INODE);
-	down(&JFS_IP(inode)->commit_sem);
+	mutex_lock(&JFS_IP(inode)->commit_mutex);
 
 	/*
-	 * Retest inode state after taking commit_sem
+	 * Retest inode state after taking commit_mutex
 	 */
 	if (inode->i_nlink && test_cflag(COMMIT_Dirty, inode))
 		rc = txCommit(tid, 1, &inode, wait ? COMMIT_SYNC : 0);
 
 	txEnd(tid);
-	up(&JFS_IP(inode)->commit_sem);
+	mutex_unlock(&JFS_IP(inode)->commit_mutex);
 	return rc;
 }
 
@@ -335,18 +336,18 @@ void jfs_truncate_nolock(struct inode *ip, loff_t length)
 		tid = txBegin(ip->i_sb, 0);
 
 		/*
-		 * The commit_sem cannot be taken before txBegin.
+		 * The commit_mutex cannot be taken before txBegin.
 		 * txBegin may block and there is a chance the inode
 		 * could be marked dirty and need to be committed
 		 * before txBegin unblocks
 		 */
-		down(&JFS_IP(ip)->commit_sem);
+		mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 		newsize = xtTruncate(tid, ip, length,
 				     COMMIT_TRUNCATE | COMMIT_PWMAP);
 		if (newsize < 0) {
 			txEnd(tid);
-			up(&JFS_IP(ip)->commit_sem);
+			mutex_unlock(&JFS_IP(ip)->commit_mutex);
 			break;
 		}
 
@@ -355,7 +356,7 @@ void jfs_truncate_nolock(struct inode *ip, loff_t length)
 
 		txCommit(tid, 1, &ip, 0);
 		txEnd(tid);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	} while (newsize > length);	/* Truncate isn't always atomic */
 }
 
diff --git a/fs/jfs/ioctl.c b/fs/jfs/ioctl.c
new file mode 100644
index 000000000000..67b3774820eb
--- /dev/null
+++ b/fs/jfs/ioctl.c
@@ -0,0 +1,107 @@
+/*
+ * linux/fs/jfs/ioctl.c
+ *
+ * Copyright (C) 2006 Herbert Poetzl
+ * adapted from Remy Card's ext2/ioctl.c
+ */
+
+#include <linux/fs.h>
+#include <linux/ext2_fs.h>
+#include <linux/ctype.h>
+#include <linux/capability.h>
+#include <linux/time.h>
+#include <asm/current.h>
+#include <asm/uaccess.h>
+
+#include "jfs_incore.h"
+#include "jfs_dinode.h"
+#include "jfs_inode.h"
+
+
+static struct {
+	long jfs_flag;
+	long ext2_flag;
+} jfs_map[] = {
+	{JFS_NOATIME_FL, EXT2_NOATIME_FL},
+	{JFS_DIRSYNC_FL, EXT2_DIRSYNC_FL},
+	{JFS_SYNC_FL, EXT2_SYNC_FL},
+	{JFS_SECRM_FL, EXT2_SECRM_FL},
+	{JFS_UNRM_FL, EXT2_UNRM_FL},
+	{JFS_APPEND_FL, EXT2_APPEND_FL},
+	{JFS_IMMUTABLE_FL, EXT2_IMMUTABLE_FL},
+	{0, 0},
+};
+
+static long jfs_map_ext2(unsigned long flags, int from)
+{
+	int index=0;
+	long mapped=0;
+
+	while (jfs_map[index].jfs_flag) {
+		if (from) {
+			if (jfs_map[index].ext2_flag & flags)
+				mapped |= jfs_map[index].jfs_flag;
+		} else {
+			if (jfs_map[index].jfs_flag & flags)
+				mapped |= jfs_map[index].ext2_flag;
+		}
+		index++;
+	}
+	return mapped;
+}
+
+
+int jfs_ioctl(struct inode * inode, struct file * filp, unsigned int cmd,
+		unsigned long arg)
+{
+	struct jfs_inode_info *jfs_inode = JFS_IP(inode);
+	unsigned int flags;
+
+	switch (cmd) {
+	case JFS_IOC_GETFLAGS:
+		flags = jfs_inode->mode2 & JFS_FL_USER_VISIBLE;
+		flags = jfs_map_ext2(flags, 0);
+		return put_user(flags, (int __user *) arg);
+	case JFS_IOC_SETFLAGS: {
+		unsigned int oldflags;
+
+		if (IS_RDONLY(inode))
+			return -EROFS;
+
+		if ((current->fsuid != inode->i_uid) && !capable(CAP_FOWNER))
+			return -EACCES;
+
+		if (get_user(flags, (int __user *) arg))
+			return -EFAULT;
+
+		flags = jfs_map_ext2(flags, 1);
+		if (!S_ISDIR(inode->i_mode))
+			flags &= ~JFS_DIRSYNC_FL;
+
+		oldflags = jfs_inode->mode2;
+
+		/*
+		 * The IMMUTABLE and APPEND_ONLY flags can only be changed by
+		 * the relevant capability.
+		 */
+		if ((oldflags & JFS_IMMUTABLE_FL) ||
+			((flags ^ oldflags) &
+			(JFS_APPEND_FL | JFS_IMMUTABLE_FL))) {
+			if (!capable(CAP_LINUX_IMMUTABLE))
+				return -EPERM;
+		}
+
+		flags = flags & JFS_FL_USER_MODIFIABLE;
+		flags |= oldflags & ~JFS_FL_USER_MODIFIABLE;
+		jfs_inode->mode2 = flags;
+
+		jfs_set_inode_flags(inode);
+		inode->i_ctime = CURRENT_TIME_SEC;
+		mark_inode_dirty(inode);
+		return 0;
+	}
+	default:
+		return -ENOTTY;
+	}
+}
+
diff --git a/fs/jfs/jfs_dinode.h b/fs/jfs/jfs_dinode.h
index 580a3258449b..9f2572aea561 100644
--- a/fs/jfs/jfs_dinode.h
+++ b/fs/jfs/jfs_dinode.h
@@ -139,13 +139,36 @@ struct dinode {
 
 /* more extended mode bits: attributes for OS/2 */
 #define IREADONLY	0x02000000	/* no write access to file */
-#define IARCHIVE	0x40000000	/* file archive bit */
-#define ISYSTEM		0x08000000	/* system file */
 #define IHIDDEN		0x04000000	/* hidden file */
-#define IRASH		0x4E000000	/* mask for changeable attributes */
-#define INEWNAME	0x80000000	/* non-8.3 filename format */
+#define ISYSTEM		0x08000000	/* system file */
+
 #define IDIRECTORY	0x20000000	/* directory (shadow of real bit) */
+#define IARCHIVE	0x40000000	/* file archive bit */
+#define INEWNAME	0x80000000	/* non-8.3 filename format */
+
+#define IRASH		0x4E000000	/* mask for changeable attributes */
 #define ATTRSHIFT	25	/* bits to shift to move attribute
 				   specification to mode position */
 
+/* extended attributes for Linux */
+
+#define JFS_NOATIME_FL		0x00080000 /* do not update atime */
+
+#define JFS_DIRSYNC_FL		0x00100000 /* dirsync behaviour */
+#define JFS_SYNC_FL		0x00200000 /* Synchronous updates */
+#define JFS_SECRM_FL		0x00400000 /* Secure deletion */
+#define JFS_UNRM_FL		0x00800000 /* allow for undelete */
+
+#define JFS_APPEND_FL		0x01000000 /* writes to file may only append */
+#define JFS_IMMUTABLE_FL	0x02000000 /* Immutable file */
+
+#define JFS_FL_USER_VISIBLE	0x03F80000
+#define JFS_FL_USER_MODIFIABLE	0x03F80000
+#define JFS_FL_INHERIT		0x03C80000
+
+/* These are identical to EXT[23]_IOC_GETFLAGS/SETFLAGS */
+#define JFS_IOC_GETFLAGS	_IOR('f', 1, long)
+#define JFS_IOC_SETFLAGS	_IOW('f', 2, long)
+
+
 #endif /*_H_JFS_DINODE */
diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c
index 2967b7393415..c161c98954e0 100644
--- a/fs/jfs/jfs_dmap.c
+++ b/fs/jfs/jfs_dmap.c
@@ -64,9 +64,9 @@
  *	to the persistent bitmaps in dmaps) is guarded by (busy) buffers.
  */
 
-#define BMAP_LOCK_INIT(bmp)	init_MUTEX(&bmp->db_bmaplock)
-#define BMAP_LOCK(bmp)		down(&bmp->db_bmaplock)
-#define BMAP_UNLOCK(bmp)	up(&bmp->db_bmaplock)
+#define BMAP_LOCK_INIT(bmp)	mutex_init(&bmp->db_bmaplock)
+#define BMAP_LOCK(bmp)		mutex_lock(&bmp->db_bmaplock)
+#define BMAP_UNLOCK(bmp)	mutex_unlock(&bmp->db_bmaplock)
 
 /*
  * forward references
@@ -125,7 +125,7 @@ static int dbGetL2AGSize(s64 nblocks);
  * into the table, with the table elements yielding the maximum
  * binary buddy of free bits within the character.
  */
-static s8 budtab[256] = {
+static const s8 budtab[256] = {
 	3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
 	2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
 	2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
@@ -532,10 +532,10 @@ dbUpdatePMap(struct inode *ipbmap,
 
 		lastlblkno = lblkno;
 
+		LOGSYNC_LOCK(log, flags);
 		if (mp->lsn != 0) {
 			/* inherit older/smaller lsn */
 			logdiff(diffp, mp->lsn, log);
-			LOGSYNC_LOCK(log, flags);
 			if (difft < diffp) {
 				mp->lsn = lsn;
 
@@ -548,20 +548,17 @@ dbUpdatePMap(struct inode *ipbmap,
 			logdiff(diffp, mp->clsn, log);
 			if (difft > diffp)
 				mp->clsn = tblk->clsn;
-			LOGSYNC_UNLOCK(log, flags);
 		} else {
 			mp->log = log;
 			mp->lsn = lsn;
 
 			/* insert bp after tblock in logsync list */
-			LOGSYNC_LOCK(log, flags);
-
 			log->count++;
 			list_add(&mp->synclist, &tblk->synclist);
 
 			mp->clsn = tblk->clsn;
-			LOGSYNC_UNLOCK(log, flags);
 		}
+		LOGSYNC_UNLOCK(log, flags);
 	}
 
 	/* write the last buffer. */
diff --git a/fs/jfs/jfs_dmap.h b/fs/jfs/jfs_dmap.h
index 32e25884e7e8..8b14cc8e0228 100644
--- a/fs/jfs/jfs_dmap.h
+++ b/fs/jfs/jfs_dmap.h
@@ -243,7 +243,7 @@ struct dbmap {
 struct bmap {
 	struct dbmap db_bmap;		/* on-disk aggregate map descriptor */
 	struct inode *db_ipbmap;	/* ptr to aggregate map incore inode */
-	struct semaphore db_bmaplock;	/* aggregate map lock */
+	struct mutex db_bmaplock;	/* aggregate map lock */
 	atomic_t db_active[MAXAG];	/* count of active, open files in AG */
 	u32 *db_DBmap;
 };
diff --git a/fs/jfs/jfs_dtree.c b/fs/jfs/jfs_dtree.c
index 404f33eae507..6c3f08319846 100644
--- a/fs/jfs/jfs_dtree.c
+++ b/fs/jfs/jfs_dtree.c
@@ -1005,6 +1005,9 @@ static int dtSplitUp(tid_t tid,
 
 		DT_PUTPAGE(smp);
 
+		if (!DO_INDEX(ip))
+			ip->i_size = xlen << sbi->l2bsize;
+
 		goto freeKeyName;
 	}
 
@@ -1055,7 +1058,9 @@ static int dtSplitUp(tid_t tid,
 				xaddr = addressPXD(pxd) + xlen;
 				dbFree(ip, xaddr, (s64) n);
 			}
-		}
+		} else if (!DO_INDEX(ip))
+			ip->i_size = lengthPXD(pxd) << sbi->l2bsize;
+
 
 	      extendOut:
 		DT_PUTPAGE(smp);
@@ -1098,6 +1103,9 @@ static int dtSplitUp(tid_t tid,
 		goto splitOut;
 	}
 
+	if (!DO_INDEX(ip))
+		ip->i_size += PSIZE;
+
 	/*
 	 * propagate up the router entry for the leaf page just split
 	 *
@@ -2424,6 +2432,9 @@ static int dtDeleteUp(tid_t tid, struct inode *ip,
 		break;
 	}
 
+	if (!DO_INDEX(ip))
+		ip->i_size -= PSIZE;
+
 	return 0;
 }
 
diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c
index 4879603daa1c..5549378358bf 100644
--- a/fs/jfs/jfs_extent.c
+++ b/fs/jfs/jfs_extent.c
@@ -94,7 +94,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, boolean_t abnr)
 	txBeginAnon(ip->i_sb);
 
 	/* Avoid race with jfs_commit_inode() */
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	/* validate extent length */
 	if (xlen > MAXXLEN)
@@ -136,14 +136,14 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, boolean_t abnr)
 	 */
 	nxlen = xlen;
 	if ((rc = extBalloc(ip, hint ? hint : INOHINT(ip), &nxlen, &nxaddr))) {
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		return (rc);
 	}
 
 	/* Allocate blocks to quota. */
 	if (DQUOT_ALLOC_BLOCK(ip, nxlen)) {
 		dbFree(ip, nxaddr, (s64) nxlen);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		return -EDQUOT;
 	}
 
@@ -165,7 +165,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, boolean_t abnr)
 	if (rc) {
 		dbFree(ip, nxaddr, nxlen);
 		DQUOT_FREE_BLOCK(ip, nxlen);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		return (rc);
 	}
 
@@ -177,7 +177,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, boolean_t abnr)
 
 	mark_inode_dirty(ip);
 
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	/*
 	 * COMMIT_SyncList flags an anonymous tlock on page that is on
 	 * sync list.
@@ -222,7 +222,7 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, boolean_t abnr)
 	/* This blocks if we are low on resources */
 	txBeginAnon(ip->i_sb);
 
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 	/* validate extent length */
 	if (nxlen > MAXXLEN)
 		nxlen = MAXXLEN;
@@ -258,7 +258,7 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, boolean_t abnr)
 	/* Allocat blocks to quota. */
 	if (DQUOT_ALLOC_BLOCK(ip, nxlen)) {
 		dbFree(ip, nxaddr, (s64) nxlen);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		return -EDQUOT;
 	}
 
@@ -338,7 +338,7 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, boolean_t abnr)
 
 	mark_inode_dirty(ip);
 exit:
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	return (rc);
 }
 #endif			/* _NOTYET */
@@ -439,12 +439,12 @@ int extRecord(struct inode *ip, xad_t * xp)
 
 	txBeginAnon(ip->i_sb);
 
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	/* update the extent */
 	rc = xtUpdate(0, ip, xp);
 
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	return rc;
 }
 
diff --git a/fs/jfs/jfs_imap.c b/fs/jfs/jfs_imap.c
index 31b4aa13dd4b..ccbe60aff83d 100644
--- a/fs/jfs/jfs_imap.c
+++ b/fs/jfs/jfs_imap.c
@@ -66,14 +66,14 @@ static HLIST_HEAD(aggregate_hash);
  * imap locks
  */
 /* iag free list lock */
-#define IAGFREE_LOCK_INIT(imap)		init_MUTEX(&imap->im_freelock)
-#define IAGFREE_LOCK(imap)		down(&imap->im_freelock)
-#define IAGFREE_UNLOCK(imap)		up(&imap->im_freelock)
+#define IAGFREE_LOCK_INIT(imap)		mutex_init(&imap->im_freelock)
+#define IAGFREE_LOCK(imap)		mutex_lock(&imap->im_freelock)
+#define IAGFREE_UNLOCK(imap)		mutex_unlock(&imap->im_freelock)
 
 /* per ag iag list locks */
-#define AG_LOCK_INIT(imap,index)	init_MUTEX(&(imap->im_aglock[index]))
-#define AG_LOCK(imap,agno)		down(&imap->im_aglock[agno])
-#define AG_UNLOCK(imap,agno)		up(&imap->im_aglock[agno])
+#define AG_LOCK_INIT(imap,index)	mutex_init(&(imap->im_aglock[index]))
+#define AG_LOCK(imap,agno)		mutex_lock(&imap->im_aglock[agno])
+#define AG_UNLOCK(imap,agno)		mutex_unlock(&imap->im_aglock[agno])
 
 /*
  * forward references
@@ -1261,7 +1261,7 @@ int diFree(struct inode *ip)
 	 * to be freed by the transaction;  
 	 */
 	tid = txBegin(ipimap->i_sb, COMMIT_FORCE);
-	down(&JFS_IP(ipimap)->commit_sem);
+	mutex_lock(&JFS_IP(ipimap)->commit_mutex);
 
 	/* acquire tlock of the iag page of the freed ixad 
 	 * to force the page NOHOMEOK (even though no data is
@@ -1294,7 +1294,7 @@ int diFree(struct inode *ip)
 	rc = txCommit(tid, 1, &iplist[0], COMMIT_FORCE);
 
 	txEnd(tid);
-	up(&JFS_IP(ipimap)->commit_sem);
+	mutex_unlock(&JFS_IP(ipimap)->commit_mutex);
 
 	/* unlock the AG inode map information */
 	AG_UNLOCK(imap, agno);
@@ -2554,13 +2554,13 @@ diNewIAG(struct inomap * imap, int *iagnop, int agno, struct metapage ** mpp)
 		 * addressing structure pointing to the new iag page;
 		 */
 		tid = txBegin(sb, COMMIT_FORCE);
-		down(&JFS_IP(ipimap)->commit_sem);
+		mutex_lock(&JFS_IP(ipimap)->commit_mutex);
 
 		/* update the inode map addressing structure to point to it */
 		if ((rc =
 		     xtInsert(tid, ipimap, 0, blkno, xlen, &xaddr, 0))) {
 			txEnd(tid);
-			up(&JFS_IP(ipimap)->commit_sem);
+			mutex_unlock(&JFS_IP(ipimap)->commit_mutex);
 			/* Free the blocks allocated for the iag since it was
 			 * not successfully added to the inode map
 			 */
@@ -2626,7 +2626,7 @@ diNewIAG(struct inomap * imap, int *iagnop, int agno, struct metapage ** mpp)
 		rc = txCommit(tid, 1, &iplist[0], COMMIT_FORCE);
 
 		txEnd(tid);
-		up(&JFS_IP(ipimap)->commit_sem);
+		mutex_unlock(&JFS_IP(ipimap)->commit_mutex);
 
 		duplicateIXtree(sb, blkno, xlen, &xaddr);
 
@@ -2844,11 +2844,11 @@ diUpdatePMap(struct inode *ipimap,
 	 */
 	lsn = tblk->lsn;
 	log = JFS_SBI(tblk->sb)->log;
+	LOGSYNC_LOCK(log, flags);
 	if (mp->lsn != 0) {
 		/* inherit older/smaller lsn */
 		logdiff(difft, lsn, log);
 		logdiff(diffp, mp->lsn, log);
-		LOGSYNC_LOCK(log, flags);
 		if (difft < diffp) {
 			mp->lsn = lsn;
 			/* move mp after tblock in logsync list */
@@ -2860,17 +2860,15 @@ diUpdatePMap(struct inode *ipimap,
 		logdiff(diffp, mp->clsn, log);
 		if (difft > diffp)
 			mp->clsn = tblk->clsn;
-		LOGSYNC_UNLOCK(log, flags);
 	} else {
 		mp->log = log;
 		mp->lsn = lsn;
 		/* insert mp after tblock in logsync list */
-		LOGSYNC_LOCK(log, flags);
 		log->count++;
 		list_add(&mp->synclist, &tblk->synclist);
 		mp->clsn = tblk->clsn;
-		LOGSYNC_UNLOCK(log, flags);
 	}
+	LOGSYNC_UNLOCK(log, flags);
 	write_metapage(mp);
 	return (0);
 }
@@ -3076,14 +3074,40 @@ static void duplicateIXtree(struct super_block *sb, s64 blkno,
 static int copy_from_dinode(struct dinode * dip, struct inode *ip)
 {
 	struct jfs_inode_info *jfs_ip = JFS_IP(ip);
+	struct jfs_sb_info *sbi = JFS_SBI(ip->i_sb);
 
 	jfs_ip->fileset = le32_to_cpu(dip->di_fileset);
 	jfs_ip->mode2 = le32_to_cpu(dip->di_mode);
 
 	ip->i_mode = le32_to_cpu(dip->di_mode) & 0xffff;
+	if (sbi->umask != -1) {
+		ip->i_mode = (ip->i_mode & ~0777) | (0777 & ~sbi->umask);
+		/* For directories, add x permission if r is allowed by umask */
+		if (S_ISDIR(ip->i_mode)) {
+			if (ip->i_mode & 0400)
+				ip->i_mode |= 0100;
+			if (ip->i_mode & 0040)
+				ip->i_mode |= 0010;
+			if (ip->i_mode & 0004)
+				ip->i_mode |= 0001;
+		}
+	}
 	ip->i_nlink = le32_to_cpu(dip->di_nlink);
-	ip->i_uid = le32_to_cpu(dip->di_uid);
-	ip->i_gid = le32_to_cpu(dip->di_gid);
+
+	jfs_ip->saved_uid = le32_to_cpu(dip->di_uid);
+	if (sbi->uid == -1)
+		ip->i_uid = jfs_ip->saved_uid;
+	else {
+		ip->i_uid = sbi->uid;
+	}
+
+	jfs_ip->saved_gid = le32_to_cpu(dip->di_gid);
+	if (sbi->gid == -1)
+		ip->i_gid = jfs_ip->saved_gid;
+	else {
+		ip->i_gid = sbi->gid;
+	}
+
 	ip->i_size = le64_to_cpu(dip->di_size);
 	ip->i_atime.tv_sec = le32_to_cpu(dip->di_atime.tv_sec);
 	ip->i_atime.tv_nsec = le32_to_cpu(dip->di_atime.tv_nsec);
@@ -3134,21 +3158,33 @@ static int copy_from_dinode(struct dinode * dip, struct inode *ip)
 static void copy_to_dinode(struct dinode * dip, struct inode *ip)
 {
 	struct jfs_inode_info *jfs_ip = JFS_IP(ip);
+	struct jfs_sb_info *sbi = JFS_SBI(ip->i_sb);
 
 	dip->di_fileset = cpu_to_le32(jfs_ip->fileset);
-	dip->di_inostamp = cpu_to_le32(JFS_SBI(ip->i_sb)->inostamp);
+	dip->di_inostamp = cpu_to_le32(sbi->inostamp);
 	dip->di_number = cpu_to_le32(ip->i_ino);
 	dip->di_gen = cpu_to_le32(ip->i_generation);
 	dip->di_size = cpu_to_le64(ip->i_size);
 	dip->di_nblocks = cpu_to_le64(PBLK2LBLK(ip->i_sb, ip->i_blocks));
 	dip->di_nlink = cpu_to_le32(ip->i_nlink);
-	dip->di_uid = cpu_to_le32(ip->i_uid);
-	dip->di_gid = cpu_to_le32(ip->i_gid);
+	if (sbi->uid == -1)
+		dip->di_uid = cpu_to_le32(ip->i_uid);
+	else
+		dip->di_uid = cpu_to_le32(jfs_ip->saved_uid);
+	if (sbi->gid == -1)
+		dip->di_gid = cpu_to_le32(ip->i_gid);
+	else
+		dip->di_gid = cpu_to_le32(jfs_ip->saved_gid);
 	/*
 	 * mode2 is only needed for storing the higher order bits.
 	 * Trust i_mode for the lower order ones
 	 */
-	dip->di_mode = cpu_to_le32((jfs_ip->mode2 & 0xffff0000) | ip->i_mode);
+	if (sbi->umask == -1)
+		dip->di_mode = cpu_to_le32((jfs_ip->mode2 & 0xffff0000) |
+					   ip->i_mode);
+	else /* Leave the original permissions alone */
+		dip->di_mode = cpu_to_le32(jfs_ip->mode2);
+
 	dip->di_atime.tv_sec = cpu_to_le32(ip->i_atime.tv_sec);
 	dip->di_atime.tv_nsec = cpu_to_le32(ip->i_atime.tv_nsec);
 	dip->di_ctime.tv_sec = cpu_to_le32(ip->i_ctime.tv_sec);
diff --git a/fs/jfs/jfs_imap.h b/fs/jfs/jfs_imap.h
index 6b59adec036a..6e24465f0f98 100644
--- a/fs/jfs/jfs_imap.h
+++ b/fs/jfs/jfs_imap.h
@@ -140,8 +140,8 @@ struct dinomap {
 struct inomap {
 	struct dinomap im_imap;		/* 4096: inode allocation control */
 	struct inode *im_ipimap;	/* 4: ptr to inode for imap   */
-	struct semaphore im_freelock;	/* 4: iag free list lock      */
-	struct semaphore im_aglock[MAXAG];	/* 512: per AG locks          */
+	struct mutex im_freelock;	/* 4: iag free list lock      */
+	struct mutex im_aglock[MAXAG];	/* 512: per AG locks          */
 	u32 *im_DBGdimap;
 	atomic_t im_numinos;	/* num of backed inodes */
 	atomic_t im_numfree;	/* num of free backed inodes */
diff --git a/fs/jfs/jfs_incore.h b/fs/jfs/jfs_incore.h
index dc21a5bd54d4..54d73716ca8c 100644
--- a/fs/jfs/jfs_incore.h
+++ b/fs/jfs/jfs_incore.h
@@ -19,6 +19,7 @@
 #ifndef _H_JFS_INCORE
 #define _H_JFS_INCORE
 
+#include <linux/mutex.h>
 #include <linux/rwsem.h>
 #include <linux/slab.h>
 #include <linux/bitops.h>
@@ -37,6 +38,8 @@
 struct jfs_inode_info {
 	int	fileset;	/* fileset number (always 16)*/
 	uint	mode2;		/* jfs-specific mode		*/
+	uint	saved_uid;	/* saved for uid mount option */
+	uint	saved_gid;	/* saved for gid mount option */
 	pxd_t   ixpxd;		/* inode extent descriptor	*/
 	dxd_t	acl;		/* dxd describing acl	*/
 	dxd_t	ea;		/* dxd describing ea	*/
@@ -62,12 +65,12 @@ struct jfs_inode_info {
 	 */
 	struct rw_semaphore rdwrlock;
 	/*
-	 * commit_sem serializes transaction processing on an inode.
+	 * commit_mutex serializes transaction processing on an inode.
 	 * It must be taken after beginning a transaction (txBegin), since
 	 * dirty inodes may be committed while a new transaction on the
 	 * inode is blocked in txBegin or TxBeginAnon
 	 */
-	struct semaphore commit_sem;
+	struct mutex commit_mutex;
 	/* xattr_sem allows us to access the xattrs without taking i_mutex */
 	struct rw_semaphore xattr_sem;
 	lid_t	xtlid;		/* lid of xtree lock on directory */
@@ -169,6 +172,9 @@ struct jfs_sb_info {
 	uint		state;		/* mount/recovery state	*/
 	unsigned long	flag;		/* mount time flags */
 	uint		p_state;	/* state prior to going no integrity */
+	uint		uid;		/* uid to override on-disk uid */
+	uint		gid;		/* gid to override on-disk gid */
+	uint		umask;		/* umask to override on-disk umask */
 };
 
 /* jfs_sb_info commit_state */
diff --git a/fs/jfs/jfs_inode.c b/fs/jfs/jfs_inode.c
index 2af5efbfd06f..495df402916d 100644
--- a/fs/jfs/jfs_inode.c
+++ b/fs/jfs/jfs_inode.c
@@ -25,6 +25,26 @@
 #include "jfs_dinode.h"
 #include "jfs_debug.h"
 
+
+void jfs_set_inode_flags(struct inode *inode)
+{
+	unsigned int flags = JFS_IP(inode)->mode2;
+
+	inode->i_flags &= ~(S_IMMUTABLE | S_APPEND |
+		S_NOATIME | S_DIRSYNC | S_SYNC);
+
+	if (flags & JFS_IMMUTABLE_FL)
+		inode->i_flags |= S_IMMUTABLE;
+	if (flags & JFS_APPEND_FL)
+		inode->i_flags |= S_APPEND;
+	if (flags & JFS_NOATIME_FL)
+		inode->i_flags |= S_NOATIME;
+	if (flags & JFS_DIRSYNC_FL)
+		inode->i_flags |= S_DIRSYNC;
+	if (flags & JFS_SYNC_FL)
+		inode->i_flags |= S_SYNC;
+}
+
 /*
  * NAME:	ialloc()
  *
@@ -63,6 +83,13 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 		inode->i_gid = current->fsgid;
 
 	/*
+	 * New inodes need to save sane values on disk when
+	 * uid & gid mount options are used
+	 */
+	jfs_inode->saved_uid = inode->i_uid;
+	jfs_inode->saved_gid = inode->i_gid;
+
+	/*
 	 * Allocate inode to quota.
 	 */
 	if (DQUOT_ALLOC_INODE(inode)) {
@@ -74,10 +101,20 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 	}
 
 	inode->i_mode = mode;
-	if (S_ISDIR(mode))
-		jfs_inode->mode2 = IDIRECTORY | mode;
-	else
-		jfs_inode->mode2 = INLINEEA | ISPARSE | mode;
+	/* inherit flags from parent */
+	jfs_inode->mode2 = JFS_IP(parent)->mode2 & JFS_FL_INHERIT;
+
+	if (S_ISDIR(mode)) {
+		jfs_inode->mode2 |= IDIRECTORY;
+		jfs_inode->mode2 &= ~JFS_DIRSYNC_FL;
+	}
+	else {
+		jfs_inode->mode2 |= INLINEEA | ISPARSE;
+		if (S_ISLNK(mode))
+			jfs_inode->mode2 &= ~(JFS_IMMUTABLE_FL|JFS_APPEND_FL);
+	}
+	jfs_inode->mode2 |= mode;
+
 	inode->i_blksize = sb->s_blocksize;
 	inode->i_blocks = 0;
 	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
@@ -98,6 +135,7 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 	jfs_inode->atlhead = 0;
 	jfs_inode->atltail = 0;
 	jfs_inode->xtlid = 0;
+	jfs_set_inode_flags(inode);
 
 	jfs_info("ialloc returns inode = 0x%p\n", inode);
 
diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h
index b54bac576cb3..095d471b9f9a 100644
--- a/fs/jfs/jfs_inode.h
+++ b/fs/jfs/jfs_inode.h
@@ -20,6 +20,8 @@
 
 extern struct inode *ialloc(struct inode *, umode_t);
 extern int jfs_fsync(struct file *, struct dentry *, int);
+extern int jfs_ioctl(struct inode *, struct file *,
+			unsigned int, unsigned long);
 extern void jfs_read_inode(struct inode *);
 extern int jfs_commit_inode(struct inode *, int);
 extern int jfs_write_inode(struct inode*, int);
@@ -29,6 +31,7 @@ extern void jfs_truncate(struct inode *);
 extern void jfs_truncate_nolock(struct inode *, loff_t);
 extern void jfs_free_zero_link(struct inode *);
 extern struct dentry *jfs_get_parent(struct dentry *dentry);
+extern void jfs_set_inode_flags(struct inode *);
 
 extern struct address_space_operations jfs_aops;
 extern struct inode_operations jfs_dir_inode_operations;
diff --git a/fs/jfs/jfs_lock.h b/fs/jfs/jfs_lock.h
index 10ad1d086685..70ac9f7d1e00 100644
--- a/fs/jfs/jfs_lock.h
+++ b/fs/jfs/jfs_lock.h
@@ -20,6 +20,7 @@
 #define _H_JFS_LOCK
 
 #include <linux/spinlock.h>
+#include <linux/mutex.h>
 #include <linux/sched.h>
 
 /*
diff --git a/fs/jfs/jfs_logmgr.c b/fs/jfs/jfs_logmgr.c
index d27bac6acaa3..0b348b13b551 100644
--- a/fs/jfs/jfs_logmgr.c
+++ b/fs/jfs/jfs_logmgr.c
@@ -64,6 +64,7 @@
 #include <linux/interrupt.h>
 #include <linux/smp_lock.h>
 #include <linux/completion.h>
+#include <linux/kthread.h>
 #include <linux/buffer_head.h>		/* for sync_blockdev() */
 #include <linux/bio.h>
 #include <linux/suspend.h>
@@ -81,15 +82,14 @@
  */
 static struct lbuf *log_redrive_list;
 static DEFINE_SPINLOCK(log_redrive_lock);
-DECLARE_WAIT_QUEUE_HEAD(jfs_IO_thread_wait);
 
 
 /*
  *	log read/write serialization (per log)
  */
-#define LOG_LOCK_INIT(log)	init_MUTEX(&(log)->loglock)
-#define LOG_LOCK(log)		down(&((log)->loglock))
-#define LOG_UNLOCK(log)		up(&((log)->loglock))
+#define LOG_LOCK_INIT(log)	mutex_init(&(log)->loglock)
+#define LOG_LOCK(log)		mutex_lock(&((log)->loglock))
+#define LOG_UNLOCK(log)		mutex_unlock(&((log)->loglock))
 
 
 /*
@@ -1105,11 +1105,10 @@ int lmLogOpen(struct super_block *sb)
 		}
 	}
 
-	if (!(log = kmalloc(sizeof(struct jfs_log), GFP_KERNEL))) {
+	if (!(log = kzalloc(sizeof(struct jfs_log), GFP_KERNEL))) {
 		up(&jfs_log_sem);
 		return -ENOMEM;
 	}
-	memset(log, 0, sizeof(struct jfs_log));
 	INIT_LIST_HEAD(&log->sb_list);
 	init_waitqueue_head(&log->syncwait);
 
@@ -1181,9 +1180,8 @@ static int open_inline_log(struct super_block *sb)
 	struct jfs_log *log;
 	int rc;
 
-	if (!(log = kmalloc(sizeof(struct jfs_log), GFP_KERNEL)))
+	if (!(log = kzalloc(sizeof(struct jfs_log), GFP_KERNEL)))
 		return -ENOMEM;
-	memset(log, 0, sizeof(struct jfs_log));
 	INIT_LIST_HEAD(&log->sb_list);
 	init_waitqueue_head(&log->syncwait);
 
@@ -1216,12 +1214,11 @@ static int open_dummy_log(struct super_block *sb)
 
 	down(&jfs_log_sem);
 	if (!dummy_log) {
-		dummy_log = kmalloc(sizeof(struct jfs_log), GFP_KERNEL);
+		dummy_log = kzalloc(sizeof(struct jfs_log), GFP_KERNEL);
 		if (!dummy_log) {
 			up(&jfs_log_sem);
 			return -ENOMEM;
 		}
-		memset(dummy_log, 0, sizeof(struct jfs_log));
 		INIT_LIST_HEAD(&dummy_log->sb_list);
 		init_waitqueue_head(&dummy_log->syncwait);
 		dummy_log->no_integrity = 1;
@@ -1980,7 +1977,7 @@ static inline void lbmRedrive(struct lbuf *bp)
 	log_redrive_list = bp;
 	spin_unlock_irqrestore(&log_redrive_lock, flags);
 
-	wake_up(&jfs_IO_thread_wait);
+	wake_up_process(jfsIOthread);
 }
 
 
@@ -2347,13 +2344,7 @@ int jfsIOWait(void *arg)
 {
 	struct lbuf *bp;
 
-	daemonize("jfsIO");
-
-	complete(&jfsIOwait);
-
 	do {
-		DECLARE_WAITQUEUE(wq, current);
-
 		spin_lock_irq(&log_redrive_lock);
 		while ((bp = log_redrive_list) != 0) {
 			log_redrive_list = bp->l_redrive_next;
@@ -2362,21 +2353,19 @@ int jfsIOWait(void *arg)
 			lbmStartIO(bp);
 			spin_lock_irq(&log_redrive_lock);
 		}
+		spin_unlock_irq(&log_redrive_lock);
+
 		if (freezing(current)) {
-			spin_unlock_irq(&log_redrive_lock);
 			refrigerator();
 		} else {
-			add_wait_queue(&jfs_IO_thread_wait, &wq);
 			set_current_state(TASK_INTERRUPTIBLE);
-			spin_unlock_irq(&log_redrive_lock);
 			schedule();
 			current->state = TASK_RUNNING;
-			remove_wait_queue(&jfs_IO_thread_wait, &wq);
 		}
-	} while (!jfs_stop_threads);
+	} while (!kthread_should_stop());
 
 	jfs_info("jfsIOWait being killed!");
-	complete_and_exit(&jfsIOwait, 0);
+	return 0;
 }
 
 /*
diff --git a/fs/jfs/jfs_logmgr.h b/fs/jfs/jfs_logmgr.h
index e4978b5b65ee..8c6909b80014 100644
--- a/fs/jfs/jfs_logmgr.h
+++ b/fs/jfs/jfs_logmgr.h
@@ -389,7 +389,7 @@ struct jfs_log {
 	int eor;		/* 4: eor of last record in eol page */
 	struct lbuf *bp;	/* 4: current log page buffer */
 
-	struct semaphore loglock;	/* 4: log write serialization lock */
+	struct mutex loglock;	/* 4: log write serialization lock */
 
 	/* syncpt */
 	int nextsync;		/* 4: bytes to write before next syncpt */
diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c
index 8a53981f9f27..5fbaeaadccd3 100644
--- a/fs/jfs/jfs_metapage.c
+++ b/fs/jfs/jfs_metapage.c
@@ -104,10 +104,9 @@ static inline int insert_metapage(struct page *page, struct metapage *mp)
 	if (PagePrivate(page))
 		a = mp_anchor(page);
 	else {
-		a = kmalloc(sizeof(struct meta_anchor), GFP_NOFS);
+		a = kzalloc(sizeof(struct meta_anchor), GFP_NOFS);
 		if (!a)
 			return -ENOMEM;
-		memset(a, 0, sizeof(struct meta_anchor));
 		set_page_private(page, (unsigned long)a);
 		SetPagePrivate(page);
 		kmap(page);
diff --git a/fs/jfs/jfs_superblock.h b/fs/jfs/jfs_superblock.h
index fcf781bf31cb..682cf1a68a18 100644
--- a/fs/jfs/jfs_superblock.h
+++ b/fs/jfs/jfs_superblock.h
@@ -113,12 +113,9 @@ extern int jfs_mount(struct super_block *);
 extern int jfs_mount_rw(struct super_block *, int);
 extern int jfs_umount(struct super_block *);
 extern int jfs_umount_rw(struct super_block *);
-
-extern int jfs_stop_threads;
-extern struct completion jfsIOwait;
-extern wait_queue_head_t jfs_IO_thread_wait;
-extern wait_queue_head_t jfs_commit_thread_wait;
-extern wait_queue_head_t jfs_sync_thread_wait;
 extern int jfs_extendfs(struct super_block *, s64, int);
 
+extern struct task_struct *jfsIOthread;
+extern struct task_struct *jfsSyncThread;
+
 #endif /*_H_JFS_SUPERBLOCK */
diff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c
index 2ddb6b892bcf..ac3d66948e8c 100644
--- a/fs/jfs/jfs_txnmgr.c
+++ b/fs/jfs/jfs_txnmgr.c
@@ -49,6 +49,7 @@
 #include <linux/suspend.h>
 #include <linux/module.h>
 #include <linux/moduleparam.h>
+#include <linux/kthread.h>
 #include "jfs_incore.h"
 #include "jfs_inode.h"
 #include "jfs_filsys.h"
@@ -121,8 +122,7 @@ static DEFINE_SPINLOCK(jfsTxnLock);
 #define LAZY_LOCK(flags)	spin_lock_irqsave(&TxAnchor.LazyLock, flags)
 #define LAZY_UNLOCK(flags) spin_unlock_irqrestore(&TxAnchor.LazyLock, flags)
 
-DECLARE_WAIT_QUEUE_HEAD(jfs_sync_thread_wait);
-DECLARE_WAIT_QUEUE_HEAD(jfs_commit_thread_wait);
+static DECLARE_WAIT_QUEUE_HEAD(jfs_commit_thread_wait);
 static int jfs_commit_thread_waking;
 
 /*
@@ -207,7 +207,7 @@ static lid_t txLockAlloc(void)
 	if ((++TxAnchor.tlocksInUse > TxLockHWM) && (jfs_tlocks_low == 0)) {
 		jfs_info("txLockAlloc tlocks low");
 		jfs_tlocks_low = 1;
-		wake_up(&jfs_sync_thread_wait);
+		wake_up_process(jfsSyncThread);
 	}
 
 	return lid;
@@ -2743,10 +2743,6 @@ int jfs_lazycommit(void *arg)
 	unsigned long flags;
 	struct jfs_sb_info *sbi;
 
-	daemonize("jfsCommit");
-
-	complete(&jfsIOwait);
-
 	do {
 		LAZY_LOCK(flags);
 		jfs_commit_thread_waking = 0;	/* OK to wake another thread */
@@ -2806,13 +2802,13 @@ int jfs_lazycommit(void *arg)
 			current->state = TASK_RUNNING;
 			remove_wait_queue(&jfs_commit_thread_wait, &wq);
 		}
-	} while (!jfs_stop_threads);
+	} while (!kthread_should_stop());
 
 	if (!list_empty(&TxAnchor.unlock_queue))
 		jfs_err("jfs_lazycommit being killed w/pending transactions!");
 	else
 		jfs_info("jfs_lazycommit being killed\n");
-	complete_and_exit(&jfsIOwait, 0);
+	return 0;
 }
 
 void txLazyUnlock(struct tblock * tblk)
@@ -2876,10 +2872,10 @@ restart:
 		 */
 		TXN_UNLOCK();
 		tid = txBegin(ip->i_sb, COMMIT_INODE | COMMIT_FORCE);
-		down(&jfs_ip->commit_sem);
+		mutex_lock(&jfs_ip->commit_mutex);
 		txCommit(tid, 1, &ip, 0);
 		txEnd(tid);
-		up(&jfs_ip->commit_sem);
+		mutex_unlock(&jfs_ip->commit_mutex);
 		/*
 		 * Just to be safe.  I don't know how
 		 * long we can run without blocking
@@ -2932,10 +2928,6 @@ int jfs_sync(void *arg)
 	int rc;
 	tid_t tid;
 
-	daemonize("jfsSync");
-
-	complete(&jfsIOwait);
-
 	do {
 		/*
 		 * write each inode on the anonymous inode list
@@ -2952,7 +2944,7 @@ int jfs_sync(void *arg)
 				 * Inode is being freed
 				 */
 				list_del_init(&jfs_ip->anon_inode_list);
-			} else if (! down_trylock(&jfs_ip->commit_sem)) {
+			} else if (! !mutex_trylock(&jfs_ip->commit_mutex)) {
 				/*
 				 * inode will be removed from anonymous list
 				 * when it is committed
@@ -2961,7 +2953,7 @@ int jfs_sync(void *arg)
 				tid = txBegin(ip->i_sb, COMMIT_INODE);
 				rc = txCommit(tid, 1, &ip, 0);
 				txEnd(tid);
-				up(&jfs_ip->commit_sem);
+				mutex_unlock(&jfs_ip->commit_mutex);
 
 				iput(ip);
 				/*
@@ -2971,7 +2963,7 @@ int jfs_sync(void *arg)
 				cond_resched();
 				TXN_LOCK();
 			} else {
-				/* We can't get the commit semaphore.  It may
+				/* We can't get the commit mutex.  It may
 				 * be held by a thread waiting for tlock's
 				 * so let's not block here.  Save it to
 				 * put back on the anon_list.
@@ -2996,19 +2988,15 @@ int jfs_sync(void *arg)
 			TXN_UNLOCK();
 			refrigerator();
 		} else {
-			DECLARE_WAITQUEUE(wq, current);
-
-			add_wait_queue(&jfs_sync_thread_wait, &wq);
 			set_current_state(TASK_INTERRUPTIBLE);
 			TXN_UNLOCK();
 			schedule();
 			current->state = TASK_RUNNING;
-			remove_wait_queue(&jfs_sync_thread_wait, &wq);
 		}
-	} while (!jfs_stop_threads);
+	} while (!kthread_should_stop());
 
 	jfs_info("jfs_sync being killed");
-	complete_and_exit(&jfsIOwait, 0);
+	return 0;
 }
 
 #if defined(CONFIG_PROC_FS) && defined(CONFIG_JFS_DEBUG)
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index 4abbe8604302..309cee575f7d 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -104,8 +104,8 @@ static int jfs_create(struct inode *dip, struct dentry *dentry, int mode,
 
 	tid = txBegin(dip->i_sb, 0);
 
-	down(&JFS_IP(dip)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dip)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	rc = jfs_init_acl(tid, ip, dip);
 	if (rc)
@@ -165,8 +165,8 @@ static int jfs_create(struct inode *dip, struct dentry *dentry, int mode,
 
       out3:
 	txEnd(tid);
-	up(&JFS_IP(dip)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dip)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	if (rc) {
 		free_ea_wmap(ip);
 		ip->i_nlink = 0;
@@ -238,8 +238,8 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, int mode)
 
 	tid = txBegin(dip->i_sb, 0);
 
-	down(&JFS_IP(dip)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dip)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	rc = jfs_init_acl(tid, ip, dip);
 	if (rc)
@@ -300,8 +300,8 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, int mode)
 
       out3:
 	txEnd(tid);
-	up(&JFS_IP(dip)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dip)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	if (rc) {
 		free_ea_wmap(ip);
 		ip->i_nlink = 0;
@@ -365,8 +365,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry)
 
 	tid = txBegin(dip->i_sb, 0);
 
-	down(&JFS_IP(dip)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dip)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	iplist[0] = dip;
 	iplist[1] = ip;
@@ -384,8 +384,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry)
 		if (rc == -EIO)
 			txAbort(tid, 1);
 		txEnd(tid);
-		up(&JFS_IP(dip)->commit_sem);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(dip)->commit_mutex);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 
 		goto out2;
 	}
@@ -422,8 +422,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry)
 
 	txEnd(tid);
 
-	up(&JFS_IP(dip)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dip)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 
 	/*
 	 * Truncating the directory index table is not guaranteed.  It
@@ -488,8 +488,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 
 	tid = txBegin(dip->i_sb, 0);
 
-	down(&JFS_IP(dip)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dip)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	iplist[0] = dip;
 	iplist[1] = ip;
@@ -503,8 +503,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 		if (rc == -EIO)
 			txAbort(tid, 1);	/* Marks FS Dirty */
 		txEnd(tid);
-		up(&JFS_IP(dip)->commit_sem);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(dip)->commit_mutex);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		IWRITE_UNLOCK(ip);
 		goto out1;
 	}
@@ -527,8 +527,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 		if ((new_size = commitZeroLink(tid, ip)) < 0) {
 			txAbort(tid, 1);	/* Marks FS Dirty */
 			txEnd(tid);
-			up(&JFS_IP(dip)->commit_sem);
-			up(&JFS_IP(ip)->commit_sem);
+			mutex_unlock(&JFS_IP(dip)->commit_mutex);
+			mutex_unlock(&JFS_IP(ip)->commit_mutex);
 			IWRITE_UNLOCK(ip);
 			rc = new_size;
 			goto out1;
@@ -556,13 +556,13 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 
 	txEnd(tid);
 
-	up(&JFS_IP(dip)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dip)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 
 
 	while (new_size && (rc == 0)) {
 		tid = txBegin(dip->i_sb, 0);
-		down(&JFS_IP(ip)->commit_sem);
+		mutex_lock(&JFS_IP(ip)->commit_mutex);
 		new_size = xtTruncate_pmap(tid, ip, new_size);
 		if (new_size < 0) {
 			txAbort(tid, 1);	/* Marks FS Dirty */
@@ -570,7 +570,7 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 		} else
 			rc = txCommit(tid, 2, &iplist[0], COMMIT_SYNC);
 		txEnd(tid);
-		up(&JFS_IP(ip)->commit_sem);
+		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	}
 
 	if (ip->i_nlink == 0)
@@ -805,8 +805,8 @@ static int jfs_link(struct dentry *old_dentry,
 
 	tid = txBegin(ip->i_sb, 0);
 
-	down(&JFS_IP(dir)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dir)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	/*
 	 * scan parent directory for entry/freespace
@@ -847,8 +847,8 @@ static int jfs_link(struct dentry *old_dentry,
       out:
 	txEnd(tid);
 
-	up(&JFS_IP(dir)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dir)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 
 	jfs_info("jfs_link: rc:%d", rc);
 	return rc;
@@ -916,8 +916,8 @@ static int jfs_symlink(struct inode *dip, struct dentry *dentry,
 
 	tid = txBegin(dip->i_sb, 0);
 
-	down(&JFS_IP(dip)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dip)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	rc = jfs_init_security(tid, ip, dip);
 	if (rc)
@@ -1037,8 +1037,8 @@ static int jfs_symlink(struct inode *dip, struct dentry *dentry,
 
       out3:
 	txEnd(tid);
-	up(&JFS_IP(dip)->commit_sem);
-	up(&JFS_IP(ip)->commit_sem);
+	mutex_unlock(&JFS_IP(dip)->commit_mutex);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
 	if (rc) {
 		free_ea_wmap(ip);
 		ip->i_nlink = 0;
@@ -1141,13 +1141,13 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	 */
 	tid = txBegin(new_dir->i_sb, 0);
 
-	down(&JFS_IP(new_dir)->commit_sem);
-	down(&JFS_IP(old_ip)->commit_sem);
+	mutex_lock(&JFS_IP(new_dir)->commit_mutex);
+	mutex_lock(&JFS_IP(old_ip)->commit_mutex);
 	if (old_dir != new_dir)
-		down(&JFS_IP(old_dir)->commit_sem);
+		mutex_lock(&JFS_IP(old_dir)->commit_mutex);
 
 	if (new_ip) {
-		down(&JFS_IP(new_ip)->commit_sem);
+		mutex_lock(&JFS_IP(new_ip)->commit_mutex);
 		/*
 		 * Change existing directory entry to new inode number
 		 */
@@ -1160,10 +1160,10 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		if (S_ISDIR(new_ip->i_mode)) {
 			new_ip->i_nlink--;
 			if (new_ip->i_nlink) {
-				up(&JFS_IP(new_dir)->commit_sem);
-				up(&JFS_IP(old_ip)->commit_sem);
+				mutex_unlock(&JFS_IP(new_dir)->commit_mutex);
+				mutex_unlock(&JFS_IP(old_ip)->commit_mutex);
 				if (old_dir != new_dir)
-					up(&JFS_IP(old_dir)->commit_sem);
+					mutex_unlock(&JFS_IP(old_dir)->commit_mutex);
 				if (!S_ISDIR(old_ip->i_mode) && new_ip)
 					IWRITE_UNLOCK(new_ip);
 				jfs_error(new_ip->i_sb,
@@ -1282,16 +1282,16 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
       out4:
 	txEnd(tid);
 
-	up(&JFS_IP(new_dir)->commit_sem);
-	up(&JFS_IP(old_ip)->commit_sem);
+	mutex_unlock(&JFS_IP(new_dir)->commit_mutex);
+	mutex_unlock(&JFS_IP(old_ip)->commit_mutex);
 	if (old_dir != new_dir)
-		up(&JFS_IP(old_dir)->commit_sem);
+		mutex_unlock(&JFS_IP(old_dir)->commit_mutex);
 	if (new_ip)
-		up(&JFS_IP(new_ip)->commit_sem);
+		mutex_unlock(&JFS_IP(new_ip)->commit_mutex);
 
 	while (new_size && (rc == 0)) {
 		tid = txBegin(new_ip->i_sb, 0);
-		down(&JFS_IP(new_ip)->commit_sem);
+		mutex_lock(&JFS_IP(new_ip)->commit_mutex);
 		new_size = xtTruncate_pmap(tid, new_ip, new_size);
 		if (new_size < 0) {
 			txAbort(tid, 1);
@@ -1299,7 +1299,7 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		} else
 			rc = txCommit(tid, 1, &new_ip, COMMIT_SYNC);
 		txEnd(tid);
-		up(&JFS_IP(new_ip)->commit_sem);
+		mutex_unlock(&JFS_IP(new_ip)->commit_mutex);
 	}
 	if (new_ip && (new_ip->i_nlink == 0))
 		set_cflag(COMMIT_Nolink, new_ip);
@@ -1361,8 +1361,8 @@ static int jfs_mknod(struct inode *dir, struct dentry *dentry,
 
 	tid = txBegin(dir->i_sb, 0);
 
-	down(&JFS_IP(dir)->commit_sem);
-	down(&JFS_IP(ip)->commit_sem);
+	mutex_lock(&JFS_IP(dir)->commit_mutex);
+	mutex_lock(&JFS_IP(ip)->commit_mutex);
 
 	rc = jfs_init_acl(tid, ip, dir);
 	if (rc)
@@ -1407,8 +1407,8 @@ static int jfs_mknod(struct inode *dir, struct dentry *dentry,
 
       out3:
 	txEnd(tid);
-	up(&JFS_IP(ip)->commit_sem);
-	up(&JFS_IP(dir)->commit_sem);
+	mutex_unlock(&JFS_IP(ip)->commit_mutex);
+	mutex_unlock(&JFS_IP(dir)->commit_mutex);
 	if (rc) {
 		free_ea_wmap(ip);
 		ip->i_nlink = 0;
@@ -1523,6 +1523,7 @@ struct file_operations jfs_dir_operations = {
 	.read		= generic_read_dir,
 	.readdir	= jfs_readdir,
 	.fsync		= jfs_fsync,
+	.ioctl		= jfs_ioctl,
 };
 
 static int jfs_ci_hash(struct dentry *dir, struct qstr *this)
diff --git a/fs/jfs/super.c b/fs/jfs/super.c
index 8d31f1336431..18f69e6aa719 100644
--- a/fs/jfs/super.c
+++ b/fs/jfs/super.c
@@ -25,6 +25,7 @@
 #include <linux/vfs.h>
 #include <linux/mount.h>
 #include <linux/moduleparam.h>
+#include <linux/kthread.h>
 #include <linux/posix_acl.h>
 #include <asm/uaccess.h>
 #include <linux/seq_file.h>
@@ -54,11 +55,9 @@ static int commit_threads = 0;
 module_param(commit_threads, int, 0);
 MODULE_PARM_DESC(commit_threads, "Number of commit threads");
 
-int jfs_stop_threads;
-static pid_t jfsIOthread;
-static pid_t jfsCommitThread[MAX_COMMIT_THREADS];
-static pid_t jfsSyncThread;
-DECLARE_COMPLETION(jfsIOwait);
+static struct task_struct *jfsCommitThread[MAX_COMMIT_THREADS];
+struct task_struct *jfsIOthread;
+struct task_struct *jfsSyncThread;
 
 #ifdef CONFIG_JFS_DEBUG
 int jfsloglevel = JFS_LOGLEVEL_WARN;
@@ -195,7 +194,7 @@ static void jfs_put_super(struct super_block *sb)
 enum {
 	Opt_integrity, Opt_nointegrity, Opt_iocharset, Opt_resize,
 	Opt_resize_nosize, Opt_errors, Opt_ignore, Opt_err, Opt_quota,
-	Opt_usrquota, Opt_grpquota
+	Opt_usrquota, Opt_grpquota, Opt_uid, Opt_gid, Opt_umask
 };
 
 static match_table_t tokens = {
@@ -209,6 +208,9 @@ static match_table_t tokens = {
 	{Opt_ignore, "quota"},
 	{Opt_usrquota, "usrquota"},
 	{Opt_grpquota, "grpquota"},
+	{Opt_uid, "uid=%u"},
+	{Opt_gid, "gid=%u"},
+	{Opt_umask, "umask=%u"},
 	{Opt_err, NULL}
 };
 
@@ -313,7 +315,29 @@ static int parse_options(char *options, struct super_block *sb, s64 *newLVSize,
 			       "JFS: quota operations not supported\n");
 			break;
 #endif
-
+		case Opt_uid:
+		{
+			char *uid = args[0].from;
+			sbi->uid = simple_strtoul(uid, &uid, 0);
+			break;
+		}
+		case Opt_gid:
+		{
+			char *gid = args[0].from;
+			sbi->gid = simple_strtoul(gid, &gid, 0);
+			break;
+		}
+		case Opt_umask:
+		{
+			char *umask = args[0].from;
+			sbi->umask = simple_strtoul(umask, &umask, 8);
+			if (sbi->umask & ~0777) {
+				printk(KERN_ERR
+				       "JFS: Invalid value of umask\n");
+				goto cleanup;
+			}
+			break;
+		}
 		default:
 			printk("jfs: Unrecognized mount option \"%s\" "
 					" or missing value\n", p);
@@ -396,12 +420,12 @@ static int jfs_fill_super(struct super_block *sb, void *data, int silent)
 	if (!new_valid_dev(sb->s_bdev->bd_dev))
 		return -EOVERFLOW;
 
-	sbi = kmalloc(sizeof (struct jfs_sb_info), GFP_KERNEL);
+	sbi = kzalloc(sizeof (struct jfs_sb_info), GFP_KERNEL);
 	if (!sbi)
 		return -ENOSPC;
-	memset(sbi, 0, sizeof (struct jfs_sb_info));
 	sb->s_fs_info = sbi;
 	sbi->sb = sb;
+	sbi->uid = sbi->gid = sbi->umask = -1;
 
 	/* initialize the mount flag and determine the default error handler */
 	flag = JFS_ERR_REMOUNT_RO;
@@ -564,10 +588,14 @@ static int jfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
 {
 	struct jfs_sb_info *sbi = JFS_SBI(vfs->mnt_sb);
 
+	if (sbi->uid != -1)
+		seq_printf(seq, ",uid=%d", sbi->uid);
+	if (sbi->gid != -1)
+		seq_printf(seq, ",gid=%d", sbi->gid);
+	if (sbi->umask != -1)
+		seq_printf(seq, ",umask=%03o", sbi->umask);
 	if (sbi->flag & JFS_NOINTEGRITY)
 		seq_puts(seq, ",nointegrity");
-	else
-		seq_puts(seq, ",integrity");
 
 #if defined(CONFIG_QUOTA)
 	if (sbi->flag & JFS_USRQUOTA)
@@ -617,7 +645,7 @@ static void init_once(void *foo, kmem_cache_t * cachep, unsigned long flags)
 		memset(jfs_ip, 0, sizeof(struct jfs_inode_info));
 		INIT_LIST_HEAD(&jfs_ip->anon_inode_list);
 		init_rwsem(&jfs_ip->rdwrlock);
-		init_MUTEX(&jfs_ip->commit_sem);
+		mutex_init(&jfs_ip->commit_mutex);
 		init_rwsem(&jfs_ip->xattr_sem);
 		spin_lock_init(&jfs_ip->ag_lock);
 		jfs_ip->active_ag = -1;
@@ -661,12 +689,12 @@ static int __init init_jfs_fs(void)
 	/*
 	 * I/O completion thread (endio)
 	 */
-	jfsIOthread = kernel_thread(jfsIOWait, NULL, CLONE_KERNEL);
-	if (jfsIOthread < 0) {
-		jfs_err("init_jfs_fs: fork failed w/rc = %d", jfsIOthread);
+	jfsIOthread = kthread_run(jfsIOWait, NULL, "jfsIO");
+	if (IS_ERR(jfsIOthread)) {
+		rc = PTR_ERR(jfsIOthread);
+		jfs_err("init_jfs_fs: fork failed w/rc = %d", rc);
 		goto end_txmngr;
 	}
-	wait_for_completion(&jfsIOwait);	/* Wait until thread starts */
 
 	if (commit_threads < 1)
 		commit_threads = num_online_cpus();
@@ -674,24 +702,21 @@ static int __init init_jfs_fs(void)
 		commit_threads = MAX_COMMIT_THREADS;
 
 	for (i = 0; i < commit_threads; i++) {
-		jfsCommitThread[i] = kernel_thread(jfs_lazycommit, NULL,
-						   CLONE_KERNEL);
-		if (jfsCommitThread[i] < 0) {
-			jfs_err("init_jfs_fs: fork failed w/rc = %d",
-				jfsCommitThread[i]);
+		jfsCommitThread[i] = kthread_run(jfs_lazycommit, NULL, "jfsCommit");
+		if (IS_ERR(jfsCommitThread[i])) {
+			rc = PTR_ERR(jfsCommitThread[i]);
+			jfs_err("init_jfs_fs: fork failed w/rc = %d", rc);
 			commit_threads = i;
 			goto kill_committask;
 		}
-		/* Wait until thread starts */
-		wait_for_completion(&jfsIOwait);
 	}
 
-	jfsSyncThread = kernel_thread(jfs_sync, NULL, CLONE_KERNEL);
-	if (jfsSyncThread < 0) {
-		jfs_err("init_jfs_fs: fork failed w/rc = %d", jfsSyncThread);
+	jfsSyncThread = kthread_run(jfs_sync, NULL, "jfsSync");
+	if (IS_ERR(jfsSyncThread)) {
+		rc = PTR_ERR(jfsSyncThread);
+		jfs_err("init_jfs_fs: fork failed w/rc = %d", rc);
 		goto kill_committask;
 	}
-	wait_for_completion(&jfsIOwait);	/* Wait until thread starts */
 
 #ifdef PROC_FS_JFS
 	jfs_proc_init();
@@ -700,13 +725,9 @@ static int __init init_jfs_fs(void)
 	return register_filesystem(&jfs_fs_type);
 
 kill_committask:
-	jfs_stop_threads = 1;
-	wake_up_all(&jfs_commit_thread_wait);
 	for (i = 0; i < commit_threads; i++)
-		wait_for_completion(&jfsIOwait);
-
-	wake_up(&jfs_IO_thread_wait);
-	wait_for_completion(&jfsIOwait);	/* Wait for thread exit */
+		kthread_stop(jfsCommitThread[i]);
+	kthread_stop(jfsIOthread);
 end_txmngr:
 	txExit();
 free_metapage:
@@ -722,16 +743,13 @@ static void __exit exit_jfs_fs(void)
 
 	jfs_info("exit_jfs_fs called");
 
-	jfs_stop_threads = 1;
 	txExit();
 	metapage_exit();
-	wake_up(&jfs_IO_thread_wait);
-	wait_for_completion(&jfsIOwait);	/* Wait until IO thread exits */
-	wake_up_all(&jfs_commit_thread_wait);
+
+	kthread_stop(jfsIOthread);
 	for (i = 0; i < commit_threads; i++)
-		wait_for_completion(&jfsIOwait);
-	wake_up(&jfs_sync_thread_wait);
-	wait_for_completion(&jfsIOwait);	/* Wait until Sync thread exits */
+		kthread_stop(jfsCommitThread[i]);
+	kthread_stop(jfsSyncThread);
 #ifdef PROC_FS_JFS
 	jfs_proc_clean();
 #endif
diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c
index f23048f9471f..9bc5b7c055ce 100644
--- a/fs/jfs/xattr.c
+++ b/fs/jfs/xattr.c
@@ -934,13 +934,13 @@ int jfs_setxattr(struct dentry *dentry, const char *name, const void *value,
 	}
 
 	tid = txBegin(inode->i_sb, 0);
-	down(&ji->commit_sem);
+	mutex_lock(&ji->commit_mutex);
 	rc = __jfs_setxattr(tid, dentry->d_inode, name, value, value_len,
 			    flags);
 	if (!rc)
 		rc = txCommit(tid, 1, &inode, 0);
 	txEnd(tid);
-	up(&ji->commit_sem);
+	mutex_unlock(&ji->commit_mutex);
 
 	return rc;
 }
@@ -1093,12 +1093,12 @@ int jfs_removexattr(struct dentry *dentry, const char *name)
 		return rc;
 
 	tid = txBegin(inode->i_sb, 0);
-	down(&ji->commit_sem);
+	mutex_lock(&ji->commit_mutex);
 	rc = __jfs_setxattr(tid, dentry->d_inode, name, NULL, 0, XATTR_REPLACE);
 	if (!rc)
 		rc = txCommit(tid, 1, &inode, 0);
 	txEnd(tid);
-	up(&ji->commit_sem);
+	mutex_unlock(&ji->commit_mutex);
 
 	return rc;
 }
diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index 220058d8616d..970b6a6aa337 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -662,12 +662,18 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl)
 	 * reclaimed while we're stuck in the unlock call. */
 	fl->fl_u.nfs_fl.flags &= ~NFS_LCK_GRANTED;
 
+	/*
+	 * Note: the server is supposed to either grant us the unlock
+	 * request, or to deny it with NLM_LCK_DENIED_GRACE_PERIOD. In either
+	 * case, we want to unlock.
+	 */
+	do_vfs_lock(fl);
+
 	if (req->a_flags & RPC_TASK_ASYNC) {
 		status = nlmclnt_async_call(req, NLMPROC_UNLOCK,
 					&nlmclnt_unlock_ops);
 		/* Hrmf... Do the unlock early since locks_remove_posix()
 		 * really expects us to free the lock synchronously */
-		do_vfs_lock(fl);
 		if (status < 0) {
 			nlmclnt_release_lockargs(req);
 			kfree(req);
@@ -680,7 +686,6 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl)
 	if (status < 0)
 		return status;
 
-	do_vfs_lock(fl);
 	if (resp->status == NLM_LCK_GRANTED)
 		return 0;
 
diff --git a/fs/namespace.c b/fs/namespace.c
index 058a44865beb..39c81a8d6316 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1338,7 +1338,7 @@ struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
 
 	new_ns = kmalloc(sizeof(struct namespace), GFP_KERNEL);
 	if (!new_ns)
-		goto out;
+		return NULL;
 
 	atomic_set(&new_ns->count, 1);
 	INIT_LIST_HEAD(&new_ns->list);
@@ -1352,7 +1352,7 @@ struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
 	if (!new_ns->root) {
 		up_write(&namespace_sem);
 		kfree(new_ns);
-		goto out;
+		return NULL;
 	}
 	spin_lock(&vfsmount_lock);
 	list_add_tail(&new_ns->list, &new_ns->root->mnt_list);
@@ -1393,7 +1393,6 @@ struct namespace *dup_namespace(struct task_struct *tsk, struct fs_struct *fs)
 	if (altrootmnt)
 		mntput(altrootmnt);
 
-out:
 	return new_ns;
 }
 
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 04ab2fc360e7..4e9b3a1b36c5 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -57,6 +57,7 @@
 #define NFSDBG_FACILITY		NFSDBG_VFS
 #define MAX_DIRECTIO_SIZE	(4096UL << PAGE_SHIFT)
 
+static void nfs_free_user_pages(struct page **pages, int npages, int do_dirty);
 static kmem_cache_t *nfs_direct_cachep;
 
 /*
@@ -107,6 +108,15 @@ nfs_get_user_pages(int rw, unsigned long user_addr, size_t size,
 					page_count, (rw == READ), 0,
 					*pages, NULL);
 		up_read(&current->mm->mmap_sem);
+		/*
+		 * If we got fewer pages than expected from get_user_pages(),
+		 * the user buffer runs off the end of a mapping; return EFAULT.
+		 */
+		if (result >= 0 && result < page_count) {
+			nfs_free_user_pages(*pages, result, 0);
+			*pages = NULL;
+			result = -EFAULT;
+		}
 	}
 	return result;
 }
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 984ca3454d04..f8c0066e02e1 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1430,7 +1430,7 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	if (status == 0)
 		status = nfs4_do_fsinfo(server, fhandle, info);
 out:
-	return status;
+	return nfs4_map_errors(status);
 }
 
 static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr)
diff --git a/fs/nfsctl.c b/fs/nfsctl.c
index 0d4cf9486068..1c72c7f85ddc 100644
--- a/fs/nfsctl.c
+++ b/fs/nfsctl.c
@@ -98,10 +98,8 @@ asmlinkage sys_nfsservctl(int cmd, struct nfsctl_arg __user *arg, void __user *r
 	if (copy_from_user(&version, &arg->ca_version, sizeof(int)))
 		return -EFAULT;
 
-	if (version != NFSCTL_VERSION) {
-		printk(KERN_WARNING "nfsd: incompatible version in syscall.\n");
+	if (version != NFSCTL_VERSION)
 		return -EINVAL;
-	}
 
 	if (cmd < 0 || cmd >= sizeof(map)/sizeof(map[0]) || !map[cmd].name)
 		return -EINVAL;
diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
index 49bd219275db..9ee956864445 100644
--- a/fs/sysfs/dir.c
+++ b/fs/sysfs/dir.c
@@ -50,6 +50,32 @@ static struct sysfs_dirent * sysfs_new_dirent(struct sysfs_dirent * parent_sd,
 	return sd;
 }
 
+/**
+ *
+ * Return -EEXIST if there is already a sysfs element with the same name for
+ * the same parent.
+ *
+ * called with parent inode's i_mutex held
+ */
+int sysfs_dirent_exist(struct sysfs_dirent *parent_sd,
+			  const unsigned char *new)
+{
+	struct sysfs_dirent * sd;
+
+	list_for_each_entry(sd, &parent_sd->s_children, s_sibling) {
+		if (sd->s_element) {
+			const unsigned char *existing = sysfs_get_name(sd);
+			if (strcmp(existing, new))
+				continue;
+			else
+				return -EEXIST;
+		}
+	}
+
+	return 0;
+}
+
+
 int sysfs_make_dirent(struct sysfs_dirent * parent_sd, struct dentry * dentry,
 			void * element, umode_t mode, int type)
 {
@@ -102,7 +128,11 @@ static int create_dir(struct kobject * k, struct dentry * p,
 	mutex_lock(&p->d_inode->i_mutex);
 	*d = lookup_one_len(n, p, strlen(n));
 	if (!IS_ERR(*d)) {
-		error = sysfs_make_dirent(p->d_fsdata, *d, k, mode, SYSFS_DIR);
+ 		if (sysfs_dirent_exist(p->d_fsdata, n))
+  			error = -EEXIST;
+  		else
+			error = sysfs_make_dirent(p->d_fsdata, *d, k, mode,
+								SYSFS_DIR);
 		if (!error) {
 			error = sysfs_create(*d, mode, init_dir);
 			if (!error) {
@@ -302,6 +332,7 @@ void sysfs_remove_dir(struct kobject * kobj)
 	 * Drop reference from dget() on entrance.
 	 */
 	dput(dentry);
+	kobj->dentry = NULL;
 }
 
 int sysfs_rename_dir(struct kobject * kobj, const char *new_name)
@@ -479,7 +510,3 @@ struct file_operations sysfs_dir_operations = {
 	.read		= generic_read_dir,
 	.readdir	= sysfs_readdir,
 };
-
-EXPORT_SYMBOL_GPL(sysfs_create_dir);
-EXPORT_SYMBOL_GPL(sysfs_remove_dir);
-EXPORT_SYMBOL_GPL(sysfs_rename_dir);
diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c
index d0e3d8495165..5e83e7246788 100644
--- a/fs/sysfs/file.c
+++ b/fs/sysfs/file.c
@@ -301,9 +301,8 @@ static int check_perm(struct inode * inode, struct file * file)
 	/* No error? Great, allocate a buffer for the file, and store it
 	 * it in file->private_data for easy access.
 	 */
-	buffer = kmalloc(sizeof(struct sysfs_buffer),GFP_KERNEL);
+	buffer = kzalloc(sizeof(struct sysfs_buffer), GFP_KERNEL);
 	if (buffer) {
-		memset(buffer,0,sizeof(struct sysfs_buffer));
 		init_MUTEX(&buffer->sem);
 		buffer->needs_read_fill = 1;
 		buffer->ops = ops;
@@ -362,10 +361,12 @@ int sysfs_add_file(struct dentry * dir, const struct attribute * attr, int type)
 {
 	struct sysfs_dirent * parent_sd = dir->d_fsdata;
 	umode_t mode = (attr->mode & S_IALLUGO) | S_IFREG;
-	int error = 0;
+	int error = -EEXIST;
 
 	mutex_lock(&dir->d_inode->i_mutex);
-	error = sysfs_make_dirent(parent_sd, NULL, (void *) attr, mode, type);
+	if (!sysfs_dirent_exist(parent_sd, attr->name))
+		error = sysfs_make_dirent(parent_sd, NULL, (void *)attr,
+					  mode, type);
 	mutex_unlock(&dir->d_inode->i_mutex);
 
 	return error;
diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c
index 689f7bcfaf30..4c29ac41ac3e 100644
--- a/fs/sysfs/inode.c
+++ b/fs/sysfs/inode.c
@@ -54,11 +54,10 @@ int sysfs_setattr(struct dentry * dentry, struct iattr * iattr)
 
 	if (!sd_iattr) {
 		/* setting attributes for the first time, allocate now */
-		sd_iattr = kmalloc(sizeof(struct iattr), GFP_KERNEL);
+		sd_iattr = kzalloc(sizeof(struct iattr), GFP_KERNEL);
 		if (!sd_iattr)
 			return -ENOMEM;
 		/* assign default attributes */
-		memset(sd_iattr, 0, sizeof(struct iattr));
 		sd_iattr->ia_mode = sd->s_mode;
 		sd_iattr->ia_uid = 0;
 		sd_iattr->ia_gid = 0;
@@ -227,12 +226,16 @@ void sysfs_drop_dentry(struct sysfs_dirent * sd, struct dentry * parent)
 void sysfs_hash_and_remove(struct dentry * dir, const char * name)
 {
 	struct sysfs_dirent * sd;
-	struct sysfs_dirent * parent_sd = dir->d_fsdata;
+	struct sysfs_dirent * parent_sd;
+
+	if (!dir)
+		return;
 
 	if (dir->d_inode == NULL)
 		/* no inode means this hasn't been made visible yet */
 		return;
 
+	parent_sd = dir->d_fsdata;
 	mutex_lock(&dir->d_inode->i_mutex);
 	list_for_each_entry(sd, &parent_sd->s_children, s_sibling) {
 		if (!sd->s_element)
diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c
index e38d6338a20d..d2eac3ceed5f 100644
--- a/fs/sysfs/symlink.c
+++ b/fs/sysfs/symlink.c
@@ -66,6 +66,7 @@ static int sysfs_add_link(struct dentry * parent, const char * name, struct kobj
 	if (!error)
 		return 0;
 
+	kobject_put(target);
 	kfree(sl->link_name);
 exit2:
 	kfree(sl);
@@ -82,12 +83,13 @@ exit1:
 int sysfs_create_link(struct kobject * kobj, struct kobject * target, const char * name)
 {
 	struct dentry * dentry = kobj->dentry;
-	int error = 0;
+	int error = -EEXIST;
 
 	BUG_ON(!kobj || !kobj->dentry || !name);
 
 	mutex_lock(&dentry->d_inode->i_mutex);
-	error = sysfs_add_link(dentry, name, target);
+	if (!sysfs_dirent_exist(dentry->d_fsdata, name))
+		error = sysfs_add_link(dentry, name, target);
 	mutex_unlock(&dentry->d_inode->i_mutex);
 	return error;
 }
diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h
index 3f8953e0e5d0..cf11d5b789d9 100644
--- a/fs/sysfs/sysfs.h
+++ b/fs/sysfs/sysfs.h
@@ -5,6 +5,7 @@ extern kmem_cache_t *sysfs_dir_cachep;
 extern struct inode * sysfs_new_inode(mode_t mode, struct sysfs_dirent *);
 extern int sysfs_create(struct dentry *, int mode, int (*init)(struct inode *));
 
+extern int sysfs_dirent_exist(struct sysfs_dirent *, const unsigned char *);
 extern int sysfs_make_dirent(struct sysfs_dirent *, struct dentry *, void *,
 				umode_t, int);