summaryrefslogtreecommitdiffstats
path: root/fs/super.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/super.c')
-rw-r--r--fs/super.c26
1 files changed, 13 insertions, 13 deletions
diff --git a/fs/super.c b/fs/super.c
index e20b558..a66f66b 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -37,6 +37,7 @@
#include <linux/writeback.h> /* for the emergency remount stuff */
#include <linux/idr.h>
#include <linux/kobject.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
@@ -55,11 +56,10 @@ DEFINE_SPINLOCK(sb_lock);
*/
static struct super_block *alloc_super(void)
{
- struct super_block *s = kmalloc(sizeof(struct super_block), GFP_USER);
+ struct super_block *s = kzalloc(sizeof(struct super_block), GFP_USER);
static struct super_operations default_op;
if (s) {
- memset(s, 0, sizeof(struct super_block));
if (security_sb_alloc(s)) {
kfree(s);
s = NULL;
@@ -76,9 +76,9 @@ static struct super_block *alloc_super(void)
down_write(&s->s_umount);
s->s_count = S_BIAS;
atomic_set(&s->s_active, 1);
- sema_init(&s->s_vfs_rename_sem,1);
- sema_init(&s->s_dquot.dqio_sem, 1);
- sema_init(&s->s_dquot.dqonoff_sem, 1);
+ mutex_init(&s->s_vfs_rename_mutex);
+ mutex_init(&s->s_dquot.dqio_mutex);
+ mutex_init(&s->s_dquot.dqonoff_mutex);
init_rwsem(&s->s_dquot.dqptr_sem);
init_waitqueue_head(&s->s_wait_unfrozen);
s->s_maxbytes = MAX_NON_LFS;
@@ -381,9 +381,9 @@ restart:
void sync_filesystems(int wait)
{
struct super_block *sb;
- static DECLARE_MUTEX(mutex);
+ static DEFINE_MUTEX(mutex);
- down(&mutex); /* Could be down_interruptible */
+ mutex_lock(&mutex); /* Could be down_interruptible */
spin_lock(&sb_lock);
list_for_each_entry(sb, &super_blocks, s_list) {
if (!sb->s_op->sync_fs)
@@ -412,7 +412,7 @@ restart:
goto restart;
}
spin_unlock(&sb_lock);
- up(&mutex);
+ mutex_unlock(&mutex);
}
/**
@@ -693,9 +693,9 @@ struct super_block *get_sb_bdev(struct file_system_type *fs_type,
* will protect the lockfs code from trying to start a snapshot
* while we are mounting
*/
- down(&bdev->bd_mount_sem);
+ mutex_lock(&bdev->bd_mount_mutex);
s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
- up(&bdev->bd_mount_sem);
+ mutex_unlock(&bdev->bd_mount_mutex);
if (IS_ERR(s))
goto out;
@@ -712,7 +712,7 @@ struct super_block *get_sb_bdev(struct file_system_type *fs_type,
s->s_flags = flags;
strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
sb_set_blocksize(s, block_size(bdev));
- error = fill_super(s, data, flags & MS_VERBOSE ? 1 : 0);
+ error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
if (error) {
up_write(&s->s_umount);
deactivate_super(s);
@@ -756,7 +756,7 @@ struct super_block *get_sb_nodev(struct file_system_type *fs_type,
s->s_flags = flags;
- error = fill_super(s, data, flags & MS_VERBOSE ? 1 : 0);
+ error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
if (error) {
up_write(&s->s_umount);
deactivate_super(s);
@@ -785,7 +785,7 @@ struct super_block *get_sb_single(struct file_system_type *fs_type,
return s;
if (!s->s_root) {
s->s_flags = flags;
- error = fill_super(s, data, flags & MS_VERBOSE ? 1 : 0);
+ error = fill_super(s, data, flags & MS_SILENT ? 1 : 0);
if (error) {
up_write(&s->s_umount);
deactivate_super(s);
OpenPOWER on IntegriCloud