From: Hawkins Jiawei yin31149@gmail.com
commit 63095f4f3af59322bea984a6ae44337439348fe0 upstream.
Kernel iterates over ATTR_RECORDs in mft record in ntfs_attr_find(). Because the ATTR_RECORDs are next to each other, kernel can get the next ATTR_RECORD from end address of current ATTR_RECORD, through current ATTR_RECORD length field.
The problem is that during iteration, when kernel calculates the end address of current ATTR_RECORD, kernel may trigger an integer overflow bug in executing `a = (ATTR_RECORD*)((u8*)a + le32_to_cpu(a->length))`. This may wrap, leading to a forever iteration on 32bit systems.
This patch solves it by adding some checks on calculating end address of current ATTR_RECORD during iteration.
Link: https://lkml.kernel.org/r/20220831160935.3409-4-yin31149@gmail.com Link: https://lore.kernel.org/all/20220827105842.GM2030@kadam/ Signed-off-by: Hawkins Jiawei yin31149@gmail.com Suggested-by: Dan Carpenter dan.carpenter@oracle.com Cc: Anton Altaparmakov anton@tuxera.com Cc: chenxiaosong (A) chenxiaosong2@huawei.com Cc: syzkaller-bugs syzkaller-bugs@googlegroups.com Signed-off-by: Andrew Morton akpm@linux-foundation.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- fs/ntfs/attrib.c | 8 ++++++++ 1 file changed, 8 insertions(+)
--- a/fs/ntfs/attrib.c +++ b/fs/ntfs/attrib.c @@ -631,6 +631,14 @@ static int ntfs_attr_find(const ATTR_TYP return -ENOENT; if (unlikely(!a->length)) break; + + /* check whether ATTR_RECORD's length wrap */ + if ((u8 *)a + le32_to_cpu(a->length) < (u8 *)a) + break; + /* check whether ATTR_RECORD's length is within bounds */ + if ((u8 *)a + le32_to_cpu(a->length) > mrec_end) + break; + if (a->type != type) continue; /*