Btrfs: check if extent buffer is aligned to sectorsize

Thanks to fuzz testing, we can pass an invalid bytenr to extent buffer
via alloc_extent_buffer().  An unaligned eb can have more pages than it
should have, which ends up extent buffer's leak or some corrupted content
in extent buffer.

This adds a warning to let us quickly know what was happening.

Now that alloc_extent_buffer() no more returns NULL, this changes its
caller and callers of its caller to match with the new error
handling.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>

6 files changed, 28 insertions, 15 deletions

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 46025688f1d0..827c949fa4bc 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -2512,6 +2512,8 @@ read_block_for_search(struct btrfs_trans_handle *trans,
 		if (!btrfs_buffer_uptodate(tmp, 0, 0))
 			ret = -EIO;
 		free_extent_buffer(tmp);
+	} else {
+		ret = PTR_ERR(tmp);
 	}
 	return ret;
 }
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 1142127f6e5e..7b5d5e8efde6 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1098,7 +1098,7 @@ void readahead_tree_block(struct btrfs_root *root, u64 bytenr)
 	struct inode *btree_inode = root->fs_info->btree_inode;
 
 	buf = btrfs_find_create_tree_block(root, bytenr);
-	if (!buf)
+	if (IS_ERR(buf))
 		return;
 	read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree,
 				 buf, 0, WAIT_NONE, btree_get_extent, 0);
@@ -1114,7 +1114,7 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr,
 	int ret;
 
 	buf = btrfs_find_create_tree_block(root, bytenr);
-	if (!buf)
+	if (IS_ERR(buf))
 		return 0;
 
 	set_bit(EXTENT_BUFFER_READAHEAD, &buf->bflags);
@@ -1172,8 +1172,8 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
 	int ret;
 
 	buf = btrfs_find_create_tree_block(root, bytenr);
-	if (!buf)
-		return ERR_PTR(-ENOMEM);
+	if (IS_ERR(buf))
+		return buf;
 
 	ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid);
 	if (ret) {
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 689d25ac6a68..5439e85c4813 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -8016,8 +8016,9 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	struct extent_buffer *buf;
 
 	buf = btrfs_find_create_tree_block(root, bytenr);
-	if (!buf)
-		return ERR_PTR(-ENOMEM);
+	if (IS_ERR(buf))
+		return buf;
+
 	btrfs_set_header_generation(buf, trans->transid);
 	btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
 	btrfs_tree_lock(buf);
@@ -8659,8 +8660,9 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	next = btrfs_find_tree_block(root->fs_info, bytenr);
 	if (!next) {
 		next = btrfs_find_create_tree_block(root, bytenr);
-		if (!next)
-			return -ENOMEM;
+		if (IS_ERR(next))
+			return PTR_ERR(next);
+
 		btrfs_set_buffer_lockdep_class(root->root_key.objectid, next,
 					       level - 1);
 		reada = 1;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index a3412d68ad37..aaee3ef55ed8 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -4892,18 +4892,25 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 	int uptodate = 1;
 	int ret;
 
+	if (!IS_ALIGNED(start, fs_info->tree_root->sectorsize)) {
+		btrfs_err(fs_info, "bad tree block start %llu", start);
+		return ERR_PTR(-EINVAL);
+	}
+
 	eb = find_extent_buffer(fs_info, start);
 	if (eb)
 		return eb;
 
 	eb = __alloc_extent_buffer(fs_info, start, len);
 	if (!eb)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	for (i = 0; i < num_pages; i++, index++) {
 		p = find_or_create_page(mapping, index, GFP_NOFS|__GFP_NOFAIL);
-		if (!p)
+		if (!p) {
+			exists = ERR_PTR(-ENOMEM);
 			goto free_eb;
+		}
 
 		spin_lock(&mapping->private_lock);
 		if (PagePrivate(p)) {
@@ -4948,8 +4955,10 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 		set_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags);
 again:
 	ret = radix_tree_preload(GFP_NOFS);
-	if (ret)
+	if (ret) {
+		exists = ERR_PTR(ret);
 		goto free_eb;
+	}
 
 	spin_lock(&fs_info->buffer_lock);
 	ret = radix_tree_insert(&fs_info->buffer_radix,
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index aa8fed11f749..8ab1dc64cbba 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2422,8 +2422,8 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 		root_owner = btrfs_header_owner(parent);
 
 		next = btrfs_find_create_tree_block(root, bytenr);
-		if (!next)
-			return -ENOMEM;
+		if (IS_ERR(next))
+			return PTR_ERR(next);
 
 		if (*level == 1) {
 			ret = wc->process_func(root, next, wc, ptr_gen);
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index fcbda4341f7d..c3a2900c6030 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -6607,8 +6607,8 @@ int btrfs_read_sys_array(struct btrfs_root *root)
 	 * overallocate but we can keep it as-is, only the first page is used.
 	 */
 	sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET);
-	if (!sb)
-		return -ENOMEM;
+	if (IS_ERR(sb))
+		return PTR_ERR(sb);
 	set_extent_buffer_uptodate(sb);
 	btrfs_set_buffer_lockdep_class(root->root_key.objectid, sb, 0);
 	/*