[Cluster-devel] [GFS2 PATCH] GFS2: attempt to access beyond end of device creating file

[Bob Peterson]

Hi,

This patch initializes the hash table cache when inodes are
created in order to prevent using a hash table pointer left over
from a previous inode. It also adds boundary checking on the
hash table.

Regards,

Bob Peterson
Red Hat File Systems

Signed-off-by: Bob Peterson <rpeterso at redhat.com> 
--
diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c
index c35573a..c2eb20f 100644
--- a/fs/gfs2/dir.c
+++ b/fs/gfs2/dir.c
@@ -740,6 +740,7 @@ static int get_leaf_nr(struct gfs2_inode *dip, u32 index,
 	hash = gfs2_dir_get_hash_table(dip);
 	if (IS_ERR(hash))
 		return PTR_ERR(hash);
+	BUG_ON(index >= (1 << dip->i_depth));
 	*leaf_out = be64_to_cpu(*(hash + index));
 	return 0;
 }
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index 6172fa7..bd2abe3 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -1576,6 +1576,7 @@ static struct inode *gfs2_alloc_inode(struct super_block *sb)
 		ip->i_flags = 0;
 		ip->i_gl = NULL;
 		ip->i_rgd = NULL;
+		ip->i_hash_cache = NULL;
 	}
 	return &ip->i_inode;
 }