ext4: Add support for 48 bit inode i_blocks.

Use the __le16 l_i_reserved1 field of the linux2 struct of ext4_inode
to represet the higher 16 bits for i_blocks. With this change max_file
size becomes (2**48 -1 )* 512 bytes.

We add a RO_COMPAT feature to the super block to indicate that inode
have i_blocks represented as a split 48 bits. Super block with this
feature set cannot be mounted read write on a kernel with CONFIG_LSF
disabled.

Super block flag EXT4_FEATURE_RO_COMPAT_HUGE_FILE

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>

3 files changed, 119 insertions, 11 deletions

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index e6634550cfc..bb89fe727bb 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2667,6 +2667,22 @@ void ext4_get_inode_flags(struct ext4_inode_info *ei)
 	if (flags & S_DIRSYNC)
 		ei->i_flags |= EXT4_DIRSYNC_FL;
 }
+static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode,
+					struct ext4_inode_info *ei)
+{
+	blkcnt_t i_blocks ;
+	struct super_block *sb = ei->vfs_inode.i_sb;
+
+	if (EXT4_HAS_RO_COMPAT_FEATURE(sb,
+				EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) {
+		/* 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);
+		return i_blocks;
+	} else {
+		return le32_to_cpu(raw_inode->i_blocks_lo);
+	}
+}
 
 void ext4_read_inode(struct inode * inode)
 {
@@ -2715,8 +2731,8 @@ void ext4_read_inode(struct inode * inode)
 		 * recovery code: that's fine, we're about to complete
 		 * the process of deleting those. */
 	}
-	inode->i_blocks = le32_to_cpu(raw_inode->i_blocks);
 	ei->i_flags = le32_to_cpu(raw_inode->i_flags);
+	inode->i_blocks = ext4_inode_blocks(raw_inode, ei);
 	ei->i_file_acl = le32_to_cpu(raw_inode->i_file_acl_lo);
 	if (EXT4_SB(inode->i_sb)->s_es->s_creator_os !=
 	    cpu_to_le32(EXT4_OS_HURD)) {
@@ -2799,6 +2815,43 @@ bad_inode:
 	return;
 }
 
+static int ext4_inode_blocks_set(handle_t *handle,
+				struct ext4_inode *raw_inode,
+				struct ext4_inode_info *ei)
+{
+	struct inode *inode = &(ei->vfs_inode);
+	u64 i_blocks = inode->i_blocks;
+	struct super_block *sb = inode->i_sb;
+	int err = 0;
+
+	if (i_blocks <= ~0U) {
+		/*
+		 * i_blocks can be represnted in a 32 bit variable
+		 * as multiple of 512 bytes
+		 */
+		raw_inode->i_blocks_lo   = cpu_to_le32((u32)i_blocks);
+		raw_inode->i_blocks_high = 0;
+	} else if (i_blocks <= 0xffffffffffffULL) {
+		/*
+		 * i_blocks can be represented in a 48 bit variable
+		 * as multiple of 512 bytes
+		 */
+		err = ext4_update_rocompat_feature(handle, sb,
+					    EXT4_FEATURE_RO_COMPAT_HUGE_FILE);
+		if (err)
+			goto  err_out;
+		/* i_block is stored in the split  48 bit fields */
+		raw_inode->i_blocks_lo   = cpu_to_le32((u32)i_blocks);
+		raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32);
+	} else {
+		ext4_error(sb, __FUNCTION__,
+				"Wrong inode i_blocks count  %llu\n",
+				(unsigned long long)inode->i_blocks);
+	}
+err_out:
+	return err;
+}
+
 /*
  * Post the struct inode info into an on-disk inode location in the
  * buffer-cache.  This gobbles the caller's reference to the
@@ -2853,7 +2906,8 @@ static int ext4_do_update_inode(handle_t *handle,
 	EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
 	EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
 
-	raw_inode->i_blocks = cpu_to_le32(inode->i_blocks);
+	if (ext4_inode_blocks_set(handle, raw_inode, ei))
+		goto out_brelse;
 	raw_inode->i_dtime = cpu_to_le32(ei->i_dtime);
 	raw_inode->i_flags = cpu_to_le32(ei->i_flags);
 	if (EXT4_SB(inode->i_sb)->s_es->s_creator_os !=
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 7be27dbe76b..2b9dc96ec43 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1603,17 +1603,50 @@ static void ext4_orphan_cleanup (struct super_block * sb,
 
 /*
  * Maximal file size.  There is a direct, and {,double-,triple-}indirect
- * block limit, and also a limit of (2^32 - 1) 512-byte sectors in i_blocks.
- * We need to be 1 filesystem block less than the 2^32 sector limit.
+ * block limit, and also a limit of (2^48 - 1) 512-byte sectors in i_blocks.
+ * We need to be 1 filesystem block less than the 2^48 sector limit.
  */
 static loff_t ext4_max_size(int bits)
 {
 	loff_t res = EXT4_NDIR_BLOCKS;
-	/* This constant is calculated to be the largest file size for a
-	 * dense, 4k-blocksize file such that the total number of
+	int meta_blocks;
+	loff_t upper_limit;
+	/* This is calculated to be the largest file size for a
+	 * dense, file such that the total number of
 	 * sectors in the file, including data and all indirect blocks,
-	 * does not exceed 2^32. */
-	const loff_t upper_limit = 0x1ff7fffd000LL;
+	 * does not exceed 2^48 -1
+	 * __u32 i_blocks_lo and _u16 i_blocks_high representing the
+	 * total number of  512 bytes blocks of the file
+	 */
+
+	if (sizeof(blkcnt_t) < sizeof(u64)) {
+		/*
+		 * CONFIG_LSF is not enabled implies the inode
+		 * i_block represent total blocks in 512 bytes
+		 * 32 == size of vfs inode i_blocks * 8
+		 */
+		upper_limit = (1LL << 32) - 1;
+
+		/* total blocks in file system block size */
+		upper_limit >>= (bits - 9);
+
+	} else {
+		/* We use 48 bit ext4_inode i_blocks */
+		upper_limit = (1LL << 48) - 1;
+
+		/* total blocks in file system block size */
+		upper_limit >>= (bits - 9);
+	}
+
+	/* indirect blocks */
+	meta_blocks = 1;
+	/* double indirect blocks */
+	meta_blocks += 1 + (1LL << (bits-2));
+	/* tripple indirect blocks */
+	meta_blocks += 1 + (1LL << (bits-2)) + (1LL << (2*(bits-2)));
+
+	upper_limit -= meta_blocks;
+	upper_limit <<= bits;
 
 	res += 1LL << (bits-2);
 	res += 1LL << (2*(bits-2));
@@ -1621,6 +1654,10 @@ static loff_t ext4_max_size(int bits)
 	res <<= bits;
 	if (res > upper_limit)
 		res = upper_limit;
+
+	if (res > MAX_LFS_FILESIZE)
+		res = MAX_LFS_FILESIZE;
+
 	return res;
 }
 
@@ -1789,6 +1826,19 @@ static int ext4_fill_super (struct super_block *sb, void *data, int silent)
 		       sb->s_id, le32_to_cpu(features));
 		goto failed_mount;
 	}
+	if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) {
+		/*
+		 * Large file size enabled file system can only be
+		 * mount if kernel is build with CONFIG_LSF
+		 */
+		if (sizeof(root->i_blocks) < sizeof(u64) &&
+				!(sb->s_flags & MS_RDONLY)) {
+			printk(KERN_ERR "EXT4-fs: %s: Filesystem with huge "
+					"files cannot be mounted read-write "
+					"without CONFIG_LSF.\n", sb->s_id);
+			goto failed_mount;
+		}
+	}
 	blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size);
 
 	if (blocksize < EXT4_MIN_BLOCK_SIZE ||
diff --git a/include/linux/ext4_fs.h b/include/linux/ext4_fs.h
index a8f3faea8ef..be25eca9c04 100644
--- a/include/linux/ext4_fs.h
+++ b/include/linux/ext4_fs.h
@@ -282,7 +282,7 @@ struct ext4_inode {
 	__le32	i_dtime;	/* Deletion Time */
 	__le16	i_gid;		/* Low 16 bits of Group Id */
 	__le16	i_links_count;	/* Links count */
-	__le32	i_blocks;	/* Blocks count */
+	__le32	i_blocks_lo;	/* Blocks count */
 	__le32	i_flags;	/* File flags */
 	union {
 		struct {
@@ -302,7 +302,7 @@ struct ext4_inode {
 	__le32	i_obso_faddr;	/* Obsoleted fragment address */
 	union {
 		struct {
-			__le16	l_i_reserved1;	/* Obsoleted fragment number/size which are removed in ext4 */
+			__le16	l_i_blocks_high; /* were l_i_reserved1 */
 			__le16	l_i_file_acl_high;
 			__le16	l_i_uid_high;	/* these 2 fields */
 			__le16	l_i_gid_high;	/* were reserved2[0] */
@@ -404,6 +404,7 @@ do {									       \
 #if defined(__KERNEL__) || defined(__linux__)
 #define i_reserved1	osd1.linux1.l_i_reserved1
 #define i_file_acl_high	osd2.linux2.l_i_file_acl_high
+#define i_blocks_high	osd2.linux2.l_i_blocks_high
 #define i_uid_low	i_uid
 #define i_gid_low	i_gid
 #define i_uid_high	osd2.linux2.l_i_uid_high
@@ -670,6 +671,7 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 #define EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER	0x0001
 #define EXT4_FEATURE_RO_COMPAT_LARGE_FILE	0x0002
 #define EXT4_FEATURE_RO_COMPAT_BTREE_DIR	0x0004
+#define EXT4_FEATURE_RO_COMPAT_HUGE_FILE        0x0008
 #define EXT4_FEATURE_RO_COMPAT_GDT_CSUM		0x0010
 #define EXT4_FEATURE_RO_COMPAT_DIR_NLINK	0x0020
 #define EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE	0x0040
@@ -681,6 +683,7 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 #define EXT4_FEATURE_INCOMPAT_META_BG		0x0010
 #define EXT4_FEATURE_INCOMPAT_EXTENTS		0x0040 /* extents support */
 #define EXT4_FEATURE_INCOMPAT_64BIT		0x0080
+#define EXT4_FEATURE_INCOMPAT_MMP               0x0100
 #define EXT4_FEATURE_INCOMPAT_FLEX_BG		0x0200
 
 #define EXT4_FEATURE_COMPAT_SUPP	EXT2_FEATURE_COMPAT_EXT_ATTR
@@ -695,7 +698,8 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 					 EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
 					 EXT4_FEATURE_RO_COMPAT_DIR_NLINK | \
 					 EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE | \
-					 EXT4_FEATURE_RO_COMPAT_BTREE_DIR)
+					 EXT4_FEATURE_RO_COMPAT_BTREE_DIR |\
+					 EXT4_FEATURE_RO_COMPAT_HUGE_FILE)
 
 /*
  * Default values for user and/or group using reserved blocks