[RFC 2/3] BKL: remove from ext2

[Jan Blunck <jblunck@suse.de>]

Remove the big kernel lock from ext2 and replace it with a per superblock
mutex.

Signed-off-by: Jan Blunck <jblunck@suse.de>
---
 fs/ext2/inode.c            |    5 +--
 fs/ext2/super.c            |   70 +++++++++++++++++++++++++++++---------------
 include/linux/ext2_fs_sb.h |    1 +
 3 files changed, 49 insertions(+), 27 deletions(-)

diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index ade6340..8b51416 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -22,7 +22,6 @@
  *  Assorted race fixes, rewrite of ext2_get_block() by Al Viro, 2000
  */
 
-#include <linux/smp_lock.h>
 #include <linux/time.h>
 #include <linux/highuid.h>
 #include <linux/pagemap.h>
@@ -1400,11 +1399,11 @@ int ext2_write_inode(struct inode *inode, int do_sync)
 			       /* If this is the first large file
 				* created, add a flag to the superblock.
 				*/
-				lock_kernel();
+				mutex_lock(&EXT2_SB(sb)->s_mutex);
 				ext2_update_dynamic_rev(sb);
 				EXT2_SET_RO_COMPAT_FEATURE(sb,
 					EXT2_FEATURE_RO_COMPAT_LARGE_FILE);
-				unlock_kernel();
+				mutex_unlock(&EXT2_SB(sb)->s_mutex);
 				ext2_write_super(sb);
 			}
 		}
diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 5af1775..87a19df 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -26,7 +26,6 @@
 #include <linux/random.h>
 #include <linux/buffer_head.h>
 #include <linux/exportfs.h>
-#include <linux/smp_lock.h>
 #include <linux/vfs.h>
 #include <linux/seq_file.h>
 #include <linux/mount.h>
@@ -43,6 +42,7 @@ static void ext2_sync_super(struct super_block *sb,
 static int ext2_remount (struct super_block * sb, int * flags, char * data);
 static int ext2_statfs (struct dentry * dentry, struct kstatfs * buf);
 static int ext2_sync_fs(struct super_block *sb, int wait);
+static int ext2_sync_fs_locked(struct super_block *sb, int wait);
 
 void ext2_error (struct super_block * sb, const char * function,
 		 const char * fmt, ...)
@@ -83,7 +83,9 @@ void ext2_warning (struct super_block * sb, const char * function,
 	printk("\n");
 	va_end(args);
 }
-
+/*
+ * This is called with sbi->s_mutex held.
+ */
 void ext2_update_dynamic_rev(struct super_block *sb)
 {
 	struct ext2_super_block *es = EXT2_SB(sb)->s_es;
@@ -109,16 +111,17 @@ void ext2_update_dynamic_rev(struct super_block *sb)
 	 */
 }
 
+/*
+ * This is called with sb->s_umount held for writing.
+ */
 static void ext2_put_super (struct super_block * sb)
 {
 	int db_count;
 	int i;
 	struct ext2_sb_info *sbi = EXT2_SB(sb);
 
-	lock_kernel();
-
-	if (sb->s_dirt)
-		ext2_write_super(sb);
+	if (sb->s_dirt && !(sb->s_flags & MS_RDONLY))
+		ext2_sync_fs_locked(sb, 1);
 
 	ext2_xattr_put_super(sb);
 	if (!(sb->s_flags & MS_RDONLY)) {
@@ -140,8 +143,6 @@ static void ext2_put_super (struct super_block * sb)
 	sb->s_fs_info = NULL;
 	kfree(sbi->s_blockgroup_lock);
 	kfree(sbi);
-
-	unlock_kernel();
 }
 
 static struct kmem_cache * ext2_inode_cachep;
@@ -207,6 +208,7 @@ static int ext2_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	struct ext2_super_block *es = sbi->s_es;
 	unsigned long def_mount_opts;
 
+	mutex_lock(&sbi->s_mutex);
 	def_mount_opts = le32_to_cpu(es->s_default_mount_opts);
 
 	if (sbi->s_sb_block != 1)
@@ -279,6 +281,7 @@ static int ext2_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (!test_opt(sb, RESERVATION))
 		seq_puts(seq, ",noreservation");
 
+	mutex_unlock(&sbi->s_mutex);
 	return 0;
 }
 
@@ -727,6 +730,9 @@ static unsigned long descriptor_loc(struct super_block *sb,
 	return ext2_group_first_block_no(sb, bg) + has_super;
 }
 
+/*
+ * This is called with sb->s_umount held for writing.
+ */
 static int ext2_fill_super(struct super_block *sb, void *data, int silent)
 {
 	struct buffer_head * bh;
@@ -745,8 +751,6 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
 	__le32 features;
 	int err;
 
-	lock_kernel();
-
 	err = -ENOMEM;
 	sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
 	if (!sbi)
@@ -762,6 +766,12 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
 	sbi->s_sb_block = sb_block;
 
 	/*
+	 * mutex for protection of modifications of the superblock while being
+	 * write out by ext2_write_super() or ext2_sync_fs().
+	 */
+	mutex_init(&sbi->s_mutex);
+
+	/*
 	 * See what the current blocksize for the device is, and
 	 * use that as the blocksize.  Otherwise (or if the blocksize
 	 * is smaller than the default) use the default.
@@ -1066,7 +1076,6 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
 		ext2_warning(sb, __func__,
 			"mounting ext3 filesystem as ext2");
 	ext2_setup_super (sb, es, sb->s_flags & MS_RDONLY);
-	unlock_kernel();
 	return 0;
 
 cantfind_ext2:
@@ -1091,7 +1100,6 @@ failed_sbi:
 	kfree(sbi->s_blockgroup_lock);
 	kfree(sbi);
 failed_unlock:
-	unlock_kernel();
 	return ret;
 }
 
@@ -1124,11 +1132,10 @@ static void ext2_sync_super(struct super_block *sb, struct ext2_super_block *es)
  * set s_state to EXT2_VALID_FS after some corrections.
  */
 
-static int ext2_sync_fs(struct super_block *sb, int wait)
+static int ext2_sync_fs_locked(struct super_block *sb, int wait)
 {
 	struct ext2_super_block *es = EXT2_SB(sb)->s_es;
 
-	lock_kernel();
 	if (es->s_state & cpu_to_le16(EXT2_VALID_FS)) {
 		ext2_debug("setting valid to 0\n");
 		es->s_state &= cpu_to_le16(~EXT2_VALID_FS);
@@ -1138,22 +1145,34 @@ static int ext2_sync_fs(struct super_block *sb, int wait)
 			cpu_to_le32(ext2_count_free_inodes(sb));
 		es->s_mtime = cpu_to_le32(get_seconds());
 		ext2_sync_super(sb, es);
-	} else {
+	} else
 		ext2_commit_super(sb, es);
-	}
-	sb->s_dirt = 0;
-	unlock_kernel();
 
+	sb->s_dirt = 0;
 	return 0;
 }
 
+static int ext2_sync_fs(struct super_block *sb, int wait)
+{
+	struct ext2_sb_info * sbi = EXT2_SB(sb);
+	int ret;
+
+	mutex_lock(&sbi->s_mutex);
+	ret = ext2_sync_fs_locked(sb, wait);
+	mutex_unlock(&sbi->s_mutex);
+	return ret;
+}
 
 void ext2_write_super(struct super_block *sb)
 {
+	struct ext2_sb_info * sbi = EXT2_SB(sb);
+
+	mutex_lock(&sbi->s_mutex);
 	if (!(sb->s_flags & MS_RDONLY))
-		ext2_sync_fs(sb, 1);
+		ext2_sync_fs_locked(sb, 1);
 	else
 		sb->s_dirt = 0;
+	mutex_unlock(&sbi->s_mutex);
 }
 
 static int ext2_remount (struct super_block * sb, int * flags, char * data)
@@ -1165,7 +1184,7 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data)
 	unsigned long old_sb_flags;
 	int err;
 
-	lock_kernel();
+	mutex_lock(&sbi->s_mutex);
 
 	/* Store the old options */
 	old_sb_flags = sb->s_flags;
@@ -1203,13 +1222,13 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data)
 		sbi->s_mount_opt |= old_mount_opt & EXT2_MOUNT_XIP;
 	}
 	if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) {
-		unlock_kernel();
+		mutex_unlock(&sbi->s_mutex);
 		return 0;
 	}
 	if (*flags & MS_RDONLY) {
 		if (le16_to_cpu(es->s_state) & EXT2_VALID_FS ||
 		    !(sbi->s_mount_state & EXT2_VALID_FS)) {
-			unlock_kernel();
+			mutex_unlock(&sbi->s_mutex);
 			return 0;
 		}
 		/*
@@ -1238,14 +1257,14 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data)
 			sb->s_flags &= ~MS_RDONLY;
 	}
 	ext2_sync_super(sb, es);
-	unlock_kernel();
+	mutex_unlock(&sbi->s_mutex);
 	return 0;
 restore_opts:
 	sbi->s_mount_opt = old_opts.s_mount_opt;
 	sbi->s_resuid = old_opts.s_resuid;
 	sbi->s_resgid = old_opts.s_resgid;
 	sb->s_flags = old_sb_flags;
-	unlock_kernel();
+	mutex_unlock(&sbi->s_mutex);
 	return err;
 }
 
@@ -1256,6 +1275,8 @@ static int ext2_statfs (struct dentry * dentry, struct kstatfs * buf)
 	struct ext2_super_block *es = sbi->s_es;
 	u64 fsid;
 
+	mutex_lock(&sbi->s_mutex);
+
 	if (test_opt (sb, MINIX_DF))
 		sbi->s_overhead_last = 0;
 	else if (sbi->s_blocks_last != le32_to_cpu(es->s_blocks_count)) {
@@ -1310,6 +1331,7 @@ static int ext2_statfs (struct dentry * dentry, struct kstatfs * buf)
 	       le64_to_cpup((void *)es->s_uuid + sizeof(u64));
 	buf->f_fsid.val[0] = fsid & 0xFFFFFFFFUL;
 	buf->f_fsid.val[1] = (fsid >> 32) & 0xFFFFFFFFUL;
+	mutex_unlock(&sbi->s_mutex);
 	return 0;
 }
 
diff --git a/include/linux/ext2_fs_sb.h b/include/linux/ext2_fs_sb.h
index 1cdb663..964b432 100644
--- a/include/linux/ext2_fs_sb.h
+++ b/include/linux/ext2_fs_sb.h
@@ -106,6 +106,7 @@ struct ext2_sb_info {
 	spinlock_t s_rsv_window_lock;
 	struct rb_root s_rsv_window_root;
 	struct ext2_reserve_window_node s_rsv_window_head;
+	struct mutex s_mutex;
 };
 
 static inline spinlock_t *
-- 
1.6.4.2