release/5.2.47/fat-don-t-allow-to-mount-if-the-FAT-length-0.patch

From 0f782a00bc446007bdb6ddf2e535f285cfa4b31a Mon Sep 17 00:00:00 2001
From: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Date: Thu, 4 Jun 2020 16:50:56 -0700
Subject: [PATCH] fat: don't allow to mount if the FAT length == 0

commit b1b65750b8db67834482f758fc385bfa7560d228 upstream.

If FAT length == 0, the image doesn't have any data. And it can be the
cause of overlapping the root dir and FAT entries.

Also Windows treats it as invalid format.

Reported-by: syzbot+6f1624f937d9d6911e2d@syzkaller.appspotmail.com
Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Marco Elver <elver@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Link: http://lkml.kernel.org/r/87r1wz8mrd.fsf@mail.parknet.co.jp
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>

diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index c3de5c0c7e7b..45b4aa8878c3 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -1514,6 +1514,12 @@ static int fat_read_bpb(struct super_block *sb, struct fat_boot_sector *b,
 		goto out;
 	}
 
+	if (bpb->fat_fat_length == 0 && bpb->fat32_length == 0) {
+		if (!silent)
+			fat_msg(sb, KERN_ERR, "bogus number of FAT sectors");
+		goto out;
+	}
+
 	error = 0;
 
 out:
-- 
2.27.0