With btrfs-corrupt-block, one can set btree node/leaf's field, if
we assign a negative value to node/leaf, we can get various hangs,
eg. if extent_root's nritems is -2ULL, then we get stuck in
btrfs_read_block_groups() because it has a while loop and
btrfs_search_slot() on extent_root will always return the first
child.
This lets us know what's happening and returns a EINVAL to callers
instead of returning the first item.
Signed-off-by: Liu Bo <bo.li....@oracle.com>
---
fs/btrfs/ctree.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index c49a500..915d224 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -1770,6 +1770,14 @@ static noinline int generic_bin_search(struct
extent_buffer *eb,
unsigned long map_len = 0;
int err;
+ if (low > high) {
+ btrfs_err(eb->fs_info,
+ "%s: low (%d) < high (%d) eb %llu owner %llu level %d",
+ __func__, low, high, eb->start,
+ btrfs_header_owner(eb), btrfs_header_level(eb));
+ return -EINVAL;
+ }
+
while (low < high) {
mid = (low + high) / 2;
offset = p + mid * item_size;
--
2.5.5
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
