All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] Btrfs: improve check_node to avoid reading corrupted nodes
@ 2016-09-15  0:23 Liu Bo
  2016-09-21 13:44 ` David Sterba
  0 siblings, 1 reply; 2+ messages in thread
From: Liu Bo @ 2016-09-15  0:23 UTC (permalink / raw)
  To: linux-btrfs; +Cc: David Sterba

We need to check items in a node to make sure that we're reading
a valid one, otherwise we could get various crashes while processing
delayed_refs.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
 fs/btrfs/disk-io.c | 32 ++++++++++++++++++++++++++++----
 1 file changed, 28 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index b401e6d..1059e5d 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -547,9 +547,10 @@ static int check_tree_block_fsid(struct btrfs_fs_info *fs_info,
 }
 
 #define CORRUPT(reason, eb, root, slot)				\
-	btrfs_crit(root->fs_info, "corrupt leaf, %s: block=%llu,"	\
-		   "root=%llu, slot=%d", reason,			\
-	       btrfs_header_bytenr(eb),	root->objectid, slot)
+	btrfs_crit(root->fs_info, "corrupt %s, %s: block=%llu,"	\
+		   " root=%llu, slot=%d",			\
+		   btrfs_header_level(eb) == 0 ? "leaf" : "node",\
+		   reason, btrfs_header_bytenr(eb), root->objectid, slot)
 
 static noinline int check_leaf(struct btrfs_root *root,
 			       struct extent_buffer *leaf)
@@ -640,6 +641,10 @@ static noinline int check_leaf(struct btrfs_root *root,
 static int check_node(struct btrfs_root *root, struct extent_buffer *node)
 {
 	unsigned long nr = btrfs_header_nritems(node);
+	struct btrfs_key key, next_key;
+	int slot;
+	u64 bytenr;
+	int ret = 0;
 
 	if (nr == 0 || nr > BTRFS_NODEPTRS_PER_BLOCK(root)) {
 		btrfs_crit(root->fs_info,
@@ -647,7 +652,26 @@ static int check_node(struct btrfs_root *root, struct extent_buffer *node)
 			   node->start, root->objectid, nr);
 		return -EIO;
 	}
-	return 0;
+
+	for (slot = 0; slot < nr - 1; slot++) {
+		bytenr = btrfs_node_blockptr(node, slot);
+		btrfs_node_key_to_cpu(node, &key, slot);
+		btrfs_node_key_to_cpu(node, &next_key, slot + 1);
+
+		if (!bytenr) {
+			CORRUPT("invalid item slot", node, root, slot);
+			ret = -EIO;
+			goto out;
+		}
+
+		if (btrfs_comp_cpu_keys(&key, &next_key) >= 0) {
+			CORRUPT("bad key order", node, root, slot);
+			ret = -EIO;
+			goto out;
+		}
+	}
+out:
+	return ret;
 }
 
 static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
-- 
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-

^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] Btrfs: improve check_node to avoid reading corrupted nodes
  2016-09-15  0:23 [PATCH] Btrfs: improve check_node to avoid reading corrupted nodes Liu Bo
@ 2016-09-21 13:44 ` David Sterba
  0 siblings, 0 replies; 2+ messages in thread
From: David Sterba @ 2016-09-21 13:44 UTC (permalink / raw)
  To: Liu Bo; +Cc: linux-btrfs, David Sterba

On Wed, Sep 14, 2016 at 05:23:24PM -0700, Liu Bo wrote:
> We need to check items in a node to make sure that we're reading
> a valid one, otherwise we could get various crashes while processing
> delayed_refs.
> 
> Signed-off-by: Liu Bo <bo.li.liu@oracle.com>

Reviewed-by: David Sterba <dsterba@suse.com>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2016-09-21 13:46 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-09-15  0:23 [PATCH] Btrfs: improve check_node to avoid reading corrupted nodes Liu Bo
2016-09-21 13:44 ` David Sterba

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.