18 files changed, 354 insertions, 154 deletions

diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c
index 8b782b0..3ee67c6 100644
--- a/fs/nilfs2/bmap.c
+++ b/fs/nilfs2/bmap.c
@@ -35,7 +35,20 @@
 
 struct inode *nilfs_bmap_get_dat(const struct nilfs_bmap *bmap)
 {
-	return nilfs_dat_inode(NILFS_I_NILFS(bmap->b_inode));
+	return NILFS_I_NILFS(bmap->b_inode)->ns_dat;
+}
+
+static int nilfs_bmap_convert_error(struct nilfs_bmap *bmap,
+				     const char *fname, int err)
+{
+	struct inode *inode = bmap->b_inode;
+
+	if (err == -EINVAL) {
+		nilfs_error(inode->i_sb, fname,
+			    "broken bmap (inode number=%lu)\n", inode->i_ino);
+		err = -EIO;
+	}
+	return err;
 }
 
 /**
@@ -66,8 +79,10 @@ int nilfs_bmap_lookup_at_level(struct nilfs_bmap *bmap, __u64 key, int level,
 
 	down_read(&bmap->b_sem);
 	ret = bmap->b_ops->bop_lookup(bmap, key, level, ptrp);
-	if (ret < 0)
+	if (ret < 0) {
+		ret = nilfs_bmap_convert_error(bmap, __func__, ret);
 		goto out;
+	}
 	if (NILFS_BMAP_USE_VBN(bmap)) {
 		ret = nilfs_dat_translate(nilfs_bmap_get_dat(bmap), *ptrp,
 					  &blocknr);
@@ -88,7 +103,8 @@ int nilfs_bmap_lookup_contig(struct nilfs_bmap *bmap, __u64 key, __u64 *ptrp,
 	down_read(&bmap->b_sem);
 	ret = bmap->b_ops->bop_lookup_contig(bmap, key, ptrp, maxblocks);
 	up_read(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 static int nilfs_bmap_do_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr)
@@ -144,7 +160,8 @@ int nilfs_bmap_insert(struct nilfs_bmap *bmap,
 	down_write(&bmap->b_sem);
 	ret = nilfs_bmap_do_insert(bmap, key, rec);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 static int nilfs_bmap_do_delete(struct nilfs_bmap *bmap, __u64 key)
@@ -180,9 +197,12 @@ int nilfs_bmap_last_key(struct nilfs_bmap *bmap, unsigned long *key)
 
 	down_read(&bmap->b_sem);
 	ret = bmap->b_ops->bop_last_key(bmap, &lastkey);
-	if (!ret)
-		*key = lastkey;
 	up_read(&bmap->b_sem);
+
+	if (ret < 0)
+		ret = nilfs_bmap_convert_error(bmap, __func__, ret);
+	else
+		*key = lastkey;
 	return ret;
 }
 
@@ -210,7 +230,8 @@ int nilfs_bmap_delete(struct nilfs_bmap *bmap, unsigned long key)
 	down_write(&bmap->b_sem);
 	ret = nilfs_bmap_do_delete(bmap, key);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 static int nilfs_bmap_do_truncate(struct nilfs_bmap *bmap, unsigned long key)
@@ -261,7 +282,8 @@ int nilfs_bmap_truncate(struct nilfs_bmap *bmap, unsigned long key)
 	down_write(&bmap->b_sem);
 	ret = nilfs_bmap_do_truncate(bmap, key);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 /**
@@ -300,7 +322,8 @@ int nilfs_bmap_propagate(struct nilfs_bmap *bmap, struct buffer_head *bh)
 	down_write(&bmap->b_sem);
 	ret = bmap->b_ops->bop_propagate(bmap, bh);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 /**
@@ -344,7 +367,8 @@ int nilfs_bmap_assign(struct nilfs_bmap *bmap,
 	down_write(&bmap->b_sem);
 	ret = bmap->b_ops->bop_assign(bmap, bh, blocknr, binfo);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 /**
@@ -373,7 +397,8 @@ int nilfs_bmap_mark(struct nilfs_bmap *bmap, __u64 key, int level)
 	down_write(&bmap->b_sem);
 	ret = bmap->b_ops->bop_mark(bmap, key, level);
 	up_write(&bmap->b_sem);
-	return ret;
+
+	return nilfs_bmap_convert_error(bmap, __func__, ret);
 }
 
 /**
diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c
index 5115814c..388e9e8 100644
--- a/fs/nilfs2/btnode.c
+++ b/fs/nilfs2/btnode.c
@@ -104,8 +104,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
 	if (pblocknr == 0) {
 		pblocknr = blocknr;
 		if (inode->i_ino != NILFS_DAT_INO) {
-			struct inode *dat =
-				nilfs_dat_inode(NILFS_I_NILFS(inode));
+			struct inode *dat = NILFS_I_NILFS(inode)->ns_dat;
 
 			/* blocknr is a virtual block number */
 			err = nilfs_dat_translate(dat, blocknr, &pblocknr);
diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index cb003c8..9d45773 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -91,7 +91,6 @@ static void nilfs_commit_chunk(struct page *page,
 			       unsigned from, unsigned to)
 {
 	struct inode *dir = mapping->host;
-	struct nilfs_sb_info *sbi = NILFS_SB(dir->i_sb);
 	loff_t pos = page_offset(page) + from;
 	unsigned len = to - from;
 	unsigned nr_dirty, copied;
@@ -103,7 +102,7 @@ static void nilfs_commit_chunk(struct page *page,
 		i_size_write(dir, pos + copied);
 	if (IS_DIRSYNC(dir))
 		nilfs_set_transaction_flag(NILFS_TI_SYNC);
-	err = nilfs_set_file_dirty(sbi, dir, nr_dirty);
+	err = nilfs_set_file_dirty(dir, nr_dirty);
 	WARN_ON(err); /* do not happen */
 	unlock_page(page);
 }
diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c
index c9a30d7..2f560c9 100644
--- a/fs/nilfs2/file.c
+++ b/fs/nilfs2/file.c
@@ -155,6 +155,7 @@ const struct inode_operations nilfs_file_inode_operations = {
 	.truncate	= nilfs_truncate,
 	.setattr	= nilfs_setattr,
 	.permission     = nilfs_permission,
+	.fiemap		= nilfs_fiemap,
 };
 
 /* end of file */
diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c
index 9f8a2da..bfc73d3 100644
--- a/fs/nilfs2/ifile.c
+++ b/fs/nilfs2/ifile.c
@@ -149,14 +149,9 @@ int nilfs_ifile_get_inode_block(struct inode *ifile, ino_t ino,
 	}
 
 	err = nilfs_palloc_get_entry_block(ifile, ino, 0, out_bh);
-	if (unlikely(err)) {
-		if (err == -EINVAL)
-			nilfs_error(sb, __func__, "ifile is broken");
-		else
-			nilfs_warning(sb, __func__,
-				      "unable to read inode: %lu",
-				      (unsigned long) ino);
-	}
+	if (unlikely(err))
+		nilfs_warning(sb, __func__, "unable to read inode: %lu",
+			      (unsigned long) ino);
 	return err;
 }
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 71d4bc8..2fd440d 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -58,7 +58,7 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff,
 	struct nilfs_inode_info *ii = NILFS_I(inode);
 	__u64 blknum = 0;
 	int err = 0, ret;
-	struct inode *dat = nilfs_dat_inode(NILFS_I_NILFS(inode));
+	struct inode *dat = NILFS_I_NILFS(inode)->ns_dat;
 	unsigned maxblocks = bh_result->b_size >> inode->i_blkbits;
 
 	down_read(&NILFS_MDT(dat)->mi_sem);
@@ -96,11 +96,6 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff,
 				       inode->i_ino,
 				       (unsigned long long)blkoff);
 				err = 0;
-			} else if (err == -EINVAL) {
-				nilfs_error(inode->i_sb, __func__,
-					    "broken bmap (inode=%lu)\n",
-					    inode->i_ino);
-				err = -EIO;
 			}
 			nilfs_transaction_abort(inode->i_sb);
 			goto out;
@@ -109,6 +104,7 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff,
 		nilfs_transaction_commit(inode->i_sb); /* never fails */
 		/* Error handling should be detailed */
 		set_buffer_new(bh_result);
+		set_buffer_delay(bh_result);
 		map_bh(bh_result, inode->i_sb, 0); /* dbn must be changed
 						      to proper value */
 	} else if (ret == -ENOENT) {
@@ -185,10 +181,9 @@ static int nilfs_set_page_dirty(struct page *page)
 
 	if (ret) {
 		struct inode *inode = page->mapping->host;
-		struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb);
 		unsigned nr_dirty = 1 << (PAGE_SHIFT - inode->i_blkbits);
 
-		nilfs_set_file_dirty(sbi, inode, nr_dirty);
+		nilfs_set_file_dirty(inode, nr_dirty);
 	}
 	return ret;
 }
@@ -229,7 +224,7 @@ static int nilfs_write_end(struct file *file, struct address_space *mapping,
 						  start + copied);
 	copied = generic_write_end(file, mapping, pos, len, copied, page,
 				   fsdata);
-	nilfs_set_file_dirty(NILFS_SB(inode->i_sb), inode, nr_dirty);
+	nilfs_set_file_dirty(inode, nr_dirty);
 	err = nilfs_transaction_commit(inode->i_sb);
 	return err ? : copied;
 }
@@ -425,13 +420,12 @@ static int __nilfs_read_inode(struct super_block *sb,
 			      struct nilfs_root *root, unsigned long ino,
 			      struct inode *inode)
 {
-	struct nilfs_sb_info *sbi = NILFS_SB(sb);
-	struct inode *dat = nilfs_dat_inode(sbi->s_nilfs);
+	struct the_nilfs *nilfs = NILFS_SB(sb)->s_nilfs;
 	struct buffer_head *bh;
 	struct nilfs_inode *raw_inode;
 	int err;
 
-	down_read(&NILFS_MDT(dat)->mi_sem);	/* XXX */
+	down_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
 	err = nilfs_ifile_get_inode_block(root->ifile, ino, &bh);
 	if (unlikely(err))
 		goto bad_inode;
@@ -461,7 +455,7 @@ static int __nilfs_read_inode(struct super_block *sb,
 	}
 	nilfs_ifile_unmap_inode(root->ifile, ino, bh);
 	brelse(bh);
-	up_read(&NILFS_MDT(dat)->mi_sem);	/* XXX */
+	up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
 	nilfs_set_inode_flags(inode);
 	return 0;
 
@@ -470,7 +464,7 @@ static int __nilfs_read_inode(struct super_block *sb,
 	brelse(bh);
 
  bad_inode:
-	up_read(&NILFS_MDT(dat)->mi_sem);	/* XXX */
+	up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
 	return err;
 }
 
@@ -629,7 +623,7 @@ static void nilfs_truncate_bmap(struct nilfs_inode_info *ii,
 
 	if (!test_bit(NILFS_I_BMAP, &ii->i_state))
 		return;
- repeat:
+repeat:
 	ret = nilfs_bmap_last_key(ii->i_bmap, &b);
 	if (ret == -ENOENT)
 		return;
@@ -646,14 +640,10 @@ static void nilfs_truncate_bmap(struct nilfs_inode_info *ii,
 		     nilfs_bmap_truncate(ii->i_bmap, b) == 0))
 		goto repeat;
 
- failed:
-	if (ret == -EINVAL)
-		nilfs_error(ii->vfs_inode.i_sb, __func__,
-			    "bmap is broken (ino=%lu)", ii->vfs_inode.i_ino);
-	else
-		nilfs_warning(ii->vfs_inode.i_sb, __func__,
-			      "failed to truncate bmap (ino=%lu, err=%d)",
-			      ii->vfs_inode.i_ino, ret);
+failed:
+	nilfs_warning(ii->vfs_inode.i_sb, __func__,
+		      "failed to truncate bmap (ino=%lu, err=%d)",
+		      ii->vfs_inode.i_ino, ret);
 }
 
 void nilfs_truncate(struct inode *inode)
@@ -682,7 +672,7 @@ void nilfs_truncate(struct inode *inode)
 		nilfs_set_transaction_flag(NILFS_TI_SYNC);
 
 	nilfs_mark_inode_dirty(inode);
-	nilfs_set_file_dirty(NILFS_SB(sb), inode, 0);
+	nilfs_set_file_dirty(inode, 0);
 	nilfs_transaction_commit(sb);
 	/* May construct a logical segment and may fail in sync mode.
 	   But truncate has no return value. */
@@ -785,20 +775,24 @@ out_err:
 	return err;
 }
 
-int nilfs_permission(struct inode *inode, int mask)
+int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	struct nilfs_root *root = NILFS_I(inode)->i_root;
+	struct nilfs_root *root;
 
+	if (flags & IPERM_FLAG_RCU)
+		return -ECHILD;
+
+	root = NILFS_I(inode)->i_root;
 	if ((mask & MAY_WRITE) && root &&
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
 
-	return generic_permission(inode, mask, NULL);
+	return generic_permission(inode, mask, flags, NULL);
 }
 
-int nilfs_load_inode_block(struct nilfs_sb_info *sbi, struct inode *inode,
-			   struct buffer_head **pbh)
+int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
 {
+	struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb);
 	struct nilfs_inode_info *ii = NILFS_I(inode);
 	int err;
 
@@ -839,9 +833,9 @@ int nilfs_inode_dirty(struct inode *inode)
 	return ret;
 }
 
-int nilfs_set_file_dirty(struct nilfs_sb_info *sbi, struct inode *inode,
-			 unsigned nr_dirty)
+int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty)
 {
+	struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb);
 	struct nilfs_inode_info *ii = NILFS_I(inode);
 
 	atomic_add(nr_dirty, &sbi->s_nilfs->ns_ndirtyblks);
@@ -874,11 +868,10 @@ int nilfs_set_file_dirty(struct nilfs_sb_info *sbi, struct inode *inode,
 
 int nilfs_mark_inode_dirty(struct inode *inode)
 {
-	struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb);
 	struct buffer_head *ibh;
 	int err;
 
-	err = nilfs_load_inode_block(sbi, inode, &ibh);
+	err = nilfs_load_inode_block(inode, &ibh);
 	if (unlikely(err)) {
 		nilfs_warning(inode->i_sb, __func__,
 			      "failed to reget inode block.\n");
@@ -920,3 +913,134 @@ void nilfs_dirty_inode(struct inode *inode)
 	nilfs_mark_inode_dirty(inode);
 	nilfs_transaction_commit(inode->i_sb); /* never fails */
 }
+
+int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		 __u64 start, __u64 len)
+{
+	struct the_nilfs *nilfs = NILFS_I_NILFS(inode);
+	__u64 logical = 0, phys = 0, size = 0;
+	__u32 flags = 0;
+	loff_t isize;
+	sector_t blkoff, end_blkoff;
+	sector_t delalloc_blkoff;
+	unsigned long delalloc_blklen;
+	unsigned int blkbits = inode->i_blkbits;
+	int ret, n;
+
+	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC);
+	if (ret)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
+
+	isize = i_size_read(inode);
+
+	blkoff = start >> blkbits;
+	end_blkoff = (start + len - 1) >> blkbits;
+
+	delalloc_blklen = nilfs_find_uncommitted_extent(inode, blkoff,
+							&delalloc_blkoff);
+
+	do {
+		__u64 blkphy;
+		unsigned int maxblocks;
+
+		if (delalloc_blklen && blkoff == delalloc_blkoff) {
+			if (size) {
+				/* End of the current extent */
+				ret = fiemap_fill_next_extent(
+					fieinfo, logical, phys, size, flags);
+				if (ret)
+					break;
+			}
+			if (blkoff > end_blkoff)
+				break;
+
+			flags = FIEMAP_EXTENT_MERGED | FIEMAP_EXTENT_DELALLOC;
+			logical = blkoff << blkbits;
+			phys = 0;
+			size = delalloc_blklen << blkbits;
+
+			blkoff = delalloc_blkoff + delalloc_blklen;
+			delalloc_blklen = nilfs_find_uncommitted_extent(
+				inode, blkoff, &delalloc_blkoff);
+			continue;
+		}
+
+		/*
+		 * Limit the number of blocks that we look up so as
+		 * not to get into the next delayed allocation extent.
+		 */
+		maxblocks = INT_MAX;
+		if (delalloc_blklen)
+			maxblocks = min_t(sector_t, delalloc_blkoff - blkoff,
+					  maxblocks);
+		blkphy = 0;
+
+		down_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
+		n = nilfs_bmap_lookup_contig(
+			NILFS_I(inode)->i_bmap, blkoff, &blkphy, maxblocks);
+		up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
+
+		if (n < 0) {
+			int past_eof;
+
+			if (unlikely(n != -ENOENT))
+				break; /* error */
+
+			/* HOLE */
+			blkoff++;
+			past_eof = ((blkoff << blkbits) >= isize);
+
+			if (size) {
+				/* End of the current extent */
+
+				if (past_eof)
+					flags |= FIEMAP_EXTENT_LAST;
+
+				ret = fiemap_fill_next_extent(
+					fieinfo, logical, phys, size, flags);
+				if (ret)
+					break;
+				size = 0;
+			}
+			if (blkoff > end_blkoff || past_eof)
+				break;
+		} else {
+			if (size) {
+				if (phys && blkphy << blkbits == phys + size) {
+					/* The current extent goes on */
+					size += n << blkbits;
+				} else {
+					/* Terminate the current extent */
+					ret = fiemap_fill_next_extent(
+						fieinfo, logical, phys, size,
+						flags);
+					if (ret || blkoff > end_blkoff)
+						break;
+
+					/* Start another extent */
+					flags = FIEMAP_EXTENT_MERGED;
+					logical = blkoff << blkbits;
+					phys = blkphy << blkbits;
+					size = n << blkbits;
+				}
+			} else {
+				/* Start a new extent */
+				flags = FIEMAP_EXTENT_MERGED;
+				logical = blkoff << blkbits;
+				phys = blkphy << blkbits;
+				size = n << blkbits;
+			}
+			blkoff += n;
+		}
+		cond_resched();
+	} while (true);
+
+	/* If ret is 1 then we just hit the end of the extent array */
+	if (ret == 1)
+		ret = 0;
+
+	mutex_unlock(&inode->i_mutex);
+	return ret;
+}
diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
index b185e93..4967389 100644
--- a/fs/nilfs2/ioctl.c
+++ b/fs/nilfs2/ioctl.c
@@ -233,7 +233,7 @@ nilfs_ioctl_do_get_vinfo(struct the_nilfs *nilfs, __u64 *posp, int flags,
 	int ret;
 
 	down_read(&nilfs->ns_segctor_sem);
-	ret = nilfs_dat_get_vinfo(nilfs_dat_inode(nilfs), buf, size, nmembs);
+	ret = nilfs_dat_get_vinfo(nilfs->ns_dat, buf, size, nmembs);
 	up_read(&nilfs->ns_segctor_sem);
 	return ret;
 }
@@ -242,8 +242,7 @@ static ssize_t
 nilfs_ioctl_do_get_bdescs(struct the_nilfs *nilfs, __u64 *posp, int flags,
 			  void *buf, size_t size, size_t nmembs)
 {
-	struct inode *dat = nilfs_dat_inode(nilfs);
-	struct nilfs_bmap *bmap = NILFS_I(dat)->i_bmap;
+	struct nilfs_bmap *bmap = NILFS_I(nilfs->ns_dat)->i_bmap;
 	struct nilfs_bdesc *bdescs = buf;
 	int ret, i;
 
@@ -421,7 +420,7 @@ static int nilfs_ioctl_free_vblocknrs(struct the_nilfs *nilfs,
 	size_t nmembs = argv->v_nmembs;
 	int ret;
 
-	ret = nilfs_dat_freev(nilfs_dat_inode(nilfs), buf, nmembs);
+	ret = nilfs_dat_freev(nilfs->ns_dat, buf, nmembs);
 
 	return (ret < 0) ? ret : nmembs;
 }
@@ -430,8 +429,7 @@ static int nilfs_ioctl_mark_blocks_dirty(struct the_nilfs *nilfs,
 					 struct nilfs_argv *argv, void *buf)
 {
 	size_t nmembs = argv->v_nmembs;
-	struct inode *dat = nilfs_dat_inode(nilfs);
-	struct nilfs_bmap *bmap = NILFS_I(dat)->i_bmap;
+	struct nilfs_bmap *bmap = NILFS_I(nilfs->ns_dat)->i_bmap;
 	struct nilfs_bdesc *bdescs = buf;
 	int ret, i;
 
@@ -450,7 +448,7 @@ static int nilfs_ioctl_mark_blocks_dirty(struct the_nilfs *nilfs,
 			/* skip dead block */
 			continue;
 		if (bdescs[i].bd_level == 0) {
-			ret = nilfs_mdt_mark_block_dirty(dat,
+			ret = nilfs_mdt_mark_block_dirty(nilfs->ns_dat,
 							 bdescs[i].bd_offset);
 			if (ret < 0) {
 				WARN_ON(ret == -ENOENT);
diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c
index 39a5b84..6a0e2a1 100644
--- a/fs/nilfs2/mdt.c
+++ b/fs/nilfs2/mdt.c
@@ -237,8 +237,6 @@ static int nilfs_mdt_read_block(struct inode *inode, unsigned long block,
  *
  * %-ENOENT - the specified block does not exist (hole block)
  *
- * %-EINVAL - bmap is broken. (the caller should call nilfs_error())
- *
  * %-EROFS - Read only filesystem (for create mode)
  */
 int nilfs_mdt_get_block(struct inode *inode, unsigned long blkoff, int create,
@@ -273,8 +271,6 @@ int nilfs_mdt_get_block(struct inode *inode, unsigned long blkoff, int create,
  * %-ENOMEM - Insufficient memory available.
  *
  * %-EIO - I/O error
- *
- * %-EINVAL - bmap is broken. (the caller should call nilfs_error())
  */
 int nilfs_mdt_delete_block(struct inode *inode, unsigned long block)
 {
@@ -350,8 +346,6 @@ int nilfs_mdt_forget_block(struct inode *inode, unsigned long block)
  * %-EIO - I/O error
  *
  * %-ENOENT - the specified block does not exist (hole block)
- *
- * %-EINVAL - bmap is broken. (the caller should call nilfs_error())
  */
 int nilfs_mdt_mark_block_dirty(struct inode *inode, unsigned long block)
 {
@@ -499,31 +493,29 @@ int nilfs_mdt_freeze_buffer(struct inode *inode, struct buffer_head *bh)
 	struct buffer_head *bh_frozen;
 	struct page *page;
 	int blkbits = inode->i_blkbits;
-	int ret = -ENOMEM;
 
 	page = grab_cache_page(&shadow->frozen_data, bh->b_page->index);
 	if (!page)
-		return ret;
+		return -ENOMEM;
 
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, 1 << blkbits, 0);
 
 	bh_frozen = nilfs_page_get_nth_block(page, bh_offset(bh) >> blkbits);
-	if (bh_frozen) {
-		if (!buffer_uptodate(bh_frozen))
-			nilfs_copy_buffer(bh_frozen, bh);
-		if (list_empty(&bh_frozen->b_assoc_buffers)) {
-			list_add_tail(&bh_frozen->b_assoc_buffers,
-				      &shadow->frozen_buffers);
-			set_buffer_nilfs_redirected(bh);
-		} else {
-			brelse(bh_frozen); /* already frozen */
-		}
-		ret = 0;
+
+	if (!buffer_uptodate(bh_frozen))
+		nilfs_copy_buffer(bh_frozen, bh);
+	if (list_empty(&bh_frozen->b_assoc_buffers)) {
+		list_add_tail(&bh_frozen->b_assoc_buffers,
+			      &shadow->frozen_buffers);
+		set_buffer_nilfs_redirected(bh);
+	} else {
+		brelse(bh_frozen); /* already frozen */
 	}
+
 	unlock_page(page);
 	page_cache_release(page);
-	return ret;
+	return 0;
 }
 
 struct buffer_head *
diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c
index 6e9557e..9803427 100644
--- a/fs/nilfs2/namei.c
+++ b/fs/nilfs2/namei.c
@@ -577,6 +577,7 @@ const struct inode_operations nilfs_dir_inode_operations = {
 	.rename		= nilfs_rename,
 	.setattr	= nilfs_setattr,
 	.permission	= nilfs_permission,
+	.fiemap		= nilfs_fiemap,
 };
 
 const struct inode_operations nilfs_special_inode_operations = {
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index f7560da..777e8fd 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -190,11 +190,6 @@ static inline int nilfs_doing_construction(void)
 	return nilfs_test_transaction_flag(NILFS_TI_WRITER);
 }
 
-static inline struct inode *nilfs_dat_inode(const struct the_nilfs *nilfs)
-{
-	return nilfs->ns_dat;
-}
-
 /*
  * function prototype
  */
@@ -256,14 +251,14 @@ extern void nilfs_update_inode(struct inode *, struct buffer_head *);
 extern void nilfs_truncate(struct inode *);
 extern void nilfs_evict_inode(struct inode *);
 extern int nilfs_setattr(struct dentry *, struct iattr *);
-int nilfs_permission(struct inode *inode, int mask);
-extern int nilfs_load_inode_block(struct nilfs_sb_info *, struct inode *,
-				  struct buffer_head **);
+int nilfs_permission(struct inode *inode, int mask, unsigned int flags);
+int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh);
 extern int nilfs_inode_dirty(struct inode *);
-extern int nilfs_set_file_dirty(struct nilfs_sb_info *, struct inode *,
-				unsigned);
+int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty);
 extern int nilfs_mark_inode_dirty(struct inode *);
 extern void nilfs_dirty_inode(struct inode *);
+int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		 __u64 start, __u64 len);
 
 /* super.c */
 extern struct inode *nilfs_alloc_inode(struct super_block *);
diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c
index a6c3c2e8..0c43241 100644
--- a/fs/nilfs2/page.c
+++ b/fs/nilfs2/page.c
@@ -491,7 +491,7 @@ unsigned nilfs_page_count_clean_buffers(struct page *page,
 	}
 	return nc;
 }
- 
+
 void nilfs_mapping_init_once(struct address_space *mapping)
 {
 	memset(mapping, 0, sizeof(*mapping));
@@ -546,3 +546,87 @@ int __nilfs_clear_page_dirty(struct page *page)
 	}
 	return TestClearPageDirty(page);
 }
+
+/**
+ * nilfs_find_uncommitted_extent - find extent of uncommitted data
+ * @inode: inode
+ * @start_blk: start block offset (in)
+ * @blkoff: start offset of the found extent (out)
+ *
+ * This function searches an extent of buffers marked "delayed" which
+ * starts from a block offset equal to or larger than @start_blk.  If
+ * such an extent was found, this will store the start offset in
+ * @blkoff and return its length in blocks.  Otherwise, zero is
+ * returned.
+ */
+unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
+					    sector_t start_blk,
+					    sector_t *blkoff)
+{
+	unsigned int i;
+	pgoff_t index;
+	unsigned int nblocks_in_page;
+	unsigned long length = 0;
+	sector_t b;
+	struct pagevec pvec;
+	struct page *page;
+
+	if (inode->i_mapping->nrpages == 0)
+		return 0;
+
+	index = start_blk >> (PAGE_CACHE_SHIFT - inode->i_blkbits);
+	nblocks_in_page = 1U << (PAGE_CACHE_SHIFT - inode->i_blkbits);
+
+	pagevec_init(&pvec, 0);
+
+repeat:
+	pvec.nr = find_get_pages_contig(inode->i_mapping, index, PAGEVEC_SIZE,
+					pvec.pages);
+	if (pvec.nr == 0)
+		return length;
+
+	if (length > 0 && pvec.pages[0]->index > index)
+		goto out;
+
+	b = pvec.pages[0]->index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
+	i = 0;
+	do {
+		page = pvec.pages[i];
+
+		lock_page(page);
+		if (page_has_buffers(page)) {
+			struct buffer_head *bh, *head;
+
+			bh = head = page_buffers(page);
+			do {
+				if (b < start_blk)
+					continue;
+				if (buffer_delay(bh)) {
+					if (length == 0)
+						*blkoff = b;
+					length++;
+				} else if (length > 0) {
+					goto out_locked;
+				}
+			} while (++b, bh = bh->b_this_page, bh != head);
+		} else {
+			if (length > 0)
+				goto out_locked;
+
+			b += nblocks_in_page;
+		}
+		unlock_page(page);
+
+	} while (++i < pagevec_count(&pvec));
+
+	index = page->index + 1;
+	pagevec_release(&pvec);
+	cond_resched();
+	goto repeat;
+
+out_locked:
+	unlock_page(page);
+out:
+	pagevec_release(&pvec);
+	return length;
+}
diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h
index fb9e8a8..622df27 100644
--- a/fs/nilfs2/page.h
+++ b/fs/nilfs2/page.h
@@ -66,6 +66,9 @@ void nilfs_mapping_init(struct address_space *mapping,
 			struct backing_dev_info *bdi,
 			const struct address_space_operations *aops);
 unsigned nilfs_page_count_clean_buffers(struct page *, unsigned, unsigned);
+unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
+					    sector_t start_blk,
+					    sector_t *blkoff);
 
 #define NILFS_PAGE_BUG(page, m, a...) \
 	do { nilfs_page_bug(page); BUG(); } while (0)
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index 5d2711c2..3dfcd3b 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -535,7 +535,7 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs,
 		if (unlikely(err))
 			goto failed_page;
 
-		err = nilfs_set_file_dirty(sbi, inode, 1);
+		err = nilfs_set_file_dirty(inode, 1);
 		if (unlikely(err))
 			goto failed_page;
 
diff --git a/fs/nilfs2/sb.h b/fs/nilfs2/sb.h
index 35a0715..7a17715 100644
--- a/fs/nilfs2/sb.h
+++ b/fs/nilfs2/sb.h
@@ -27,14 +27,6 @@
 #include <linux/types.h>
 #include <linux/fs.h>
 
-/*
- * Mount options
- */
-struct nilfs_mount_options {
-	unsigned long mount_opt;
-	__u64 snapshot_cno;
-};
-
 struct the_nilfs;
 struct nilfs_sc_info;
 
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 687d090..55ebae5 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -504,17 +504,6 @@ static int nilfs_segctor_add_file_block(struct nilfs_sc_info *sci,
 	return err;
 }
 
-static int nilfs_handle_bmap_error(int err, const char *fname,
-				   struct inode *inode, struct super_block *sb)
-{
-	if (err == -EINVAL) {
-		nilfs_error(sb, fname, "broken bmap (inode=%lu)\n",
-			    inode->i_ino);
-		err = -EIO;
-	}
-	return err;
-}
-
 /*
  * Callback functions that enumerate, mark, and collect dirty blocks
  */
@@ -524,9 +513,8 @@ static int nilfs_collect_file_data(struct nilfs_sc_info *sci,
 	int err;
 
 	err = nilfs_bmap_propagate(NILFS_I(inode)->i_bmap, bh);
-	if (unlikely(err < 0))
-		return nilfs_handle_bmap_error(err, __func__, inode,
-					       sci->sc_super);
+	if (err < 0)
+		return err;
 
 	err = nilfs_segctor_add_file_block(sci, bh, inode,
 					   sizeof(struct nilfs_binfo_v));
@@ -539,13 +527,7 @@ static int nilfs_collect_file_node(struct nilfs_sc_info *sci,
 				   struct buffer_head *bh,
 				   struct inode *inode)
 {
-	int err;
-
-	err = nilfs_bmap_propagate(NILFS_I(inode)->i_bmap, bh);
-	if (unlikely(err < 0))
-		return nilfs_handle_bmap_error(err, __func__, inode,
-					       sci->sc_super);
-	return 0;
+	return nilfs_bmap_propagate(NILFS_I(inode)->i_bmap, bh);
 }
 
 static int nilfs_collect_file_bmap(struct nilfs_sc_info *sci,
@@ -588,9 +570,8 @@ static int nilfs_collect_dat_data(struct nilfs_sc_info *sci,
 	int err;
 
 	err = nilfs_bmap_propagate(NILFS_I(inode)->i_bmap, bh);
-	if (unlikely(err < 0))
-		return nilfs_handle_bmap_error(err, __func__, inode,
-					       sci->sc_super);
+	if (err < 0)
+		return err;
 
 	err = nilfs_segctor_add_file_block(sci, bh, inode, sizeof(__le64));
 	if (!err)
@@ -776,9 +757,8 @@ static int nilfs_test_metadata_dirty(struct the_nilfs *nilfs,
 		ret++;
 	if (nilfs_mdt_fetch_dirty(nilfs->ns_sufile))
 		ret++;
-	if (ret || nilfs_doing_gc())
-		if (nilfs_mdt_fetch_dirty(nilfs_dat_inode(nilfs)))
-			ret++;
+	if ((ret || nilfs_doing_gc()) && nilfs_mdt_fetch_dirty(nilfs->ns_dat))
+		ret++;
 	return ret;
 }
 
@@ -814,7 +794,7 @@ static void nilfs_segctor_clear_metadata_dirty(struct nilfs_sc_info *sci)
 	nilfs_mdt_clear_dirty(sci->sc_root->ifile);
 	nilfs_mdt_clear_dirty(nilfs->ns_cpfile);
 	nilfs_mdt_clear_dirty(nilfs->ns_sufile);
-	nilfs_mdt_clear_dirty(nilfs_dat_inode(nilfs));
+	nilfs_mdt_clear_dirty(nilfs->ns_dat);
 }
 
 static int nilfs_segctor_create_checkpoint(struct nilfs_sc_info *sci)
@@ -923,7 +903,7 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci,
 			      nilfs->ns_nongc_ctime : sci->sc_seg_ctime);
 	raw_sr->sr_flags = 0;
 
-	nilfs_write_inode_common(nilfs_dat_inode(nilfs), (void *)raw_sr +
+	nilfs_write_inode_common(nilfs->ns_dat, (void *)raw_sr +
 				 NILFS_SR_DAT_OFFSET(isz), 1);
 	nilfs_write_inode_common(nilfs->ns_cpfile, (void *)raw_sr +
 				 NILFS_SR_CPFILE_OFFSET(isz), 1);
@@ -1179,7 +1159,7 @@ static int nilfs_segctor_collect_blocks(struct nilfs_sc_info *sci, int mode)
 		sci->sc_stage.scnt++;  /* Fall through */
 	case NILFS_ST_DAT:
  dat_stage:
-		err = nilfs_segctor_scan_file(sci, nilfs_dat_inode(nilfs),
+		err = nilfs_segctor_scan_file(sci, nilfs->ns_dat,
 					      &nilfs_sc_dat_ops);
 		if (unlikely(err))
 			break;
@@ -1563,7 +1543,6 @@ nilfs_segctor_update_payload_blocknr(struct nilfs_sc_info *sci,
 	return 0;
 
  failed_bmap:
-	err = nilfs_handle_bmap_error(err, __func__, inode, sci->sc_super);
 	return err;
 }
 
@@ -1783,6 +1762,7 @@ static void nilfs_clear_copied_buffers(struct list_head *list, int err)
 				if (!err) {
 					set_buffer_uptodate(bh);
 					clear_buffer_dirty(bh);
+					clear_buffer_delay(bh);
 					clear_buffer_nilfs_volatile(bh);
 				}
 				brelse(bh); /* for b_assoc_buffers */
@@ -1909,6 +1889,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci)
 				    b_assoc_buffers) {
 			set_buffer_uptodate(bh);
 			clear_buffer_dirty(bh);
+			clear_buffer_delay(bh);
 			clear_buffer_nilfs_volatile(bh);
 			clear_buffer_nilfs_redirected(bh);
 			if (bh == segbuf->sb_super_root) {
diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
index f804d41..70dfdd5 100644
--- a/fs/nilfs2/super.c
+++ b/fs/nilfs2/super.c
@@ -47,7 +47,6 @@
 #include <linux/crc32.h>
 #include <linux/vfs.h>
 #include <linux/writeback.h>
-#include <linux/kobject.h>
 #include <linux/seq_file.h>
 #include <linux/mount.h>
 #include "nilfs.h"
@@ -111,12 +110,17 @@ void nilfs_error(struct super_block *sb, const char *function,
 		 const char *fmt, ...)
 {
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
+	struct va_format vaf;
 	va_list args;
 
 	va_start(args, fmt);
-	printk(KERN_CRIT "NILFS error (device %s): %s: ", sb->s_id, function);
-	vprintk(fmt, args);
-	printk("\n");
+
+	vaf.fmt = fmt;
+	vaf.va = &args;
+
+	printk(KERN_CRIT "NILFS error (device %s): %s: %pV\n",
+	       sb->s_id, function, &vaf);
+
 	va_end(args);
 
 	if (!(sb->s_flags & MS_RDONLY)) {
@@ -136,13 +140,17 @@ void nilfs_error(struct super_block *sb, const char *function,
 void nilfs_warning(struct super_block *sb, const char *function,
 		   const char *fmt, ...)
 {
+	struct va_format vaf;
 	va_list args;
 
 	va_start(args, fmt);
-	printk(KERN_WARNING "NILFS warning (device %s): %s: ",
-	       sb->s_id, function);
-	vprintk(fmt, args);
-	printk("\n");
+
+	vaf.fmt = fmt;
+	vaf.va = &args;
+
+	printk(KERN_WARNING "NILFS warning (device %s): %s: %pV\n",
+	       sb->s_id, function, &vaf);
+
 	va_end(args);
 }
 
@@ -162,10 +170,13 @@ struct inode *nilfs_alloc_inode(struct super_block *sb)
 	return &ii->vfs_inode;
 }
 
-void nilfs_destroy_inode(struct inode *inode)
+static void nilfs_i_callback(struct rcu_head *head)
 {
+	struct inode *inode = container_of(head, struct inode, i_rcu);
 	struct nilfs_mdt_info *mdi = NILFS_MDT(inode);
 
+	INIT_LIST_HEAD(&inode->i_dentry);
+
 	if (mdi) {
 		kfree(mdi->mi_bgl); /* kfree(NULL) is safe */
 		kfree(mdi);
@@ -173,6 +184,11 @@ void nilfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(nilfs_inode_cachep, NILFS_I(inode));
 }
 
+void nilfs_destroy_inode(struct inode *inode)
+{
+	call_rcu(&inode->i_rcu, nilfs_i_callback);
+}
+
 static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag)
 {
 	struct the_nilfs *nilfs = sbi->s_nilfs;
@@ -838,7 +854,7 @@ static int nilfs_attach_snapshot(struct super_block *s, __u64 cno,
 
 static int nilfs_tree_was_touched(struct dentry *root_dentry)
 {
-	return atomic_read(&root_dentry->d_count) > 1;
+	return root_dentry->d_count > 1;
 }
 
 /**
@@ -1002,11 +1018,11 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data)
 	struct nilfs_sb_info *sbi = NILFS_SB(sb);
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	unsigned long old_sb_flags;
-	struct nilfs_mount_options old_opts;
+	unsigned long old_mount_opt;
 	int err;
 
 	old_sb_flags = sb->s_flags;
-	old_opts.mount_opt = sbi->s_mount_opt;
+	old_mount_opt = sbi->s_mount_opt;
 
 	if (!parse_options(data, sb, 1)) {
 		err = -EINVAL;
@@ -1075,7 +1091,7 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data)
 
  restore_opts:
 	sb->s_flags = old_sb_flags;
-	sbi->s_mount_opt = old_opts.mount_opt;
+	sbi->s_mount_opt = old_mount_opt;
 	return err;
 }
 
diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c
index 0254be2..ad4ac60 100644
--- a/fs/nilfs2/the_nilfs.c
+++ b/fs/nilfs2/the_nilfs.c
@@ -329,7 +329,6 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi)
 	printk(KERN_INFO "NILFS: recovery complete.\n");
 
  skip_recovery:
-	set_nilfs_loaded(nilfs);
 	nilfs_clear_recovery_info(&ri);
 	sbi->s_super->s_flags = s_flags;
 	return 0;
@@ -651,12 +650,11 @@ int nilfs_discard_segments(struct the_nilfs *nilfs, __u64 *segnump,
 
 int nilfs_count_free_blocks(struct the_nilfs *nilfs, sector_t *nblocks)
 {
-	struct inode *dat = nilfs_dat_inode(nilfs);
 	unsigned long ncleansegs;
 
-	down_read(&NILFS_MDT(dat)->mi_sem);	/* XXX */
+	down_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
 	ncleansegs = nilfs_sufile_get_ncleansegs(nilfs->ns_sufile);
-	up_read(&NILFS_MDT(dat)->mi_sem);	/* XXX */
+	up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem);
 	*nblocks = (sector_t)ncleansegs * nilfs->ns_blocks_per_segment;
 	return 0;
 }
diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h
index 69226e1..fd85e4c 100644
--- a/fs/nilfs2/the_nilfs.h
+++ b/fs/nilfs2/the_nilfs.h
@@ -36,8 +36,6 @@
 /* the_nilfs struct */
 enum {
 	THE_NILFS_INIT = 0,     /* Information from super_block is set */
-	THE_NILFS_LOADED,       /* Roll-back/roll-forward has done and
-				   the latest checkpoint was loaded */
 	THE_NILFS_DISCONTINUED,	/* 'next' pointer chain has broken */
 	THE_NILFS_GC_RUNNING,	/* gc process is running */
 	THE_NILFS_SB_DIRTY,	/* super block is dirty */
@@ -178,7 +176,6 @@ static inline int nilfs_##name(struct the_nilfs *nilfs)			\
 }
 
 THE_NILFS_FNS(INIT, init)
-THE_NILFS_FNS(LOADED, loaded)
 THE_NILFS_FNS(DISCONTINUED, discontinued)
 THE_NILFS_FNS(GC_RUNNING, gc_running)
 THE_NILFS_FNS(SB_DIRTY, sb_dirty)