Re: [PATCH next] fs: fix oob in do_handle_open

[Jeff Layton <jlayton@kernel.org>]

On Wed, 2024-04-03 at 14:54 +0800, Edward Adam Davis wrote:
> [Syzbot reported]
> BUG: KASAN: slab-out-of-bounds in instrument_copy_from_user_before include/linux/instrumented.h:129 [inline]
> BUG: KASAN: slab-out-of-bounds in _copy_from_user+0x7b/0xe0 lib/usercopy.c:22
> Write of size 48 at addr ffff88802b8cbc88 by task syz-executor333/5090
> 
> CPU: 0 PID: 5090 Comm: syz-executor333 Not tainted 6.9.0-rc2-next-20240402-syzkaller #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/27/2024
> Call Trace:
>  <TASK>
>  __dump_stack lib/dump_stack.c:88 [inline]
>  dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114
>  print_address_description mm/kasan/report.c:377 [inline]
>  print_report+0x169/0x550 mm/kasan/report.c:488
>  kasan_report+0x143/0x180 mm/kasan/report.c:601
>  kasan_check_range+0x282/0x290 mm/kasan/generic.c:189
>  instrument_copy_from_user_before include/linux/instrumented.h:129 [inline]
>  _copy_from_user+0x7b/0xe0 lib/usercopy.c:22
>  copy_from_user include/linux/uaccess.h:183 [inline]
>  handle_to_path fs/fhandle.c:203 [inline]
>  do_handle_open+0x204/0x660 fs/fhandle.c:226
>  do_syscall_64+0xfb/0x240
>  entry_SYSCALL_64_after_hwframe+0x72/0x7a
> [Fix] 
> When copying data to f_handle, the length of the copied data should not include
> the length of "struct file_handle".
> 
> Reported-by: syzbot+4139435cb1b34cf759c2@syzkaller.appspotmail.com
> Signed-off-by: Edward Adam Davis <eadavis@qq.com>
> ---
>  fs/fhandle.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/fhandle.c b/fs/fhandle.c
> index 53ed54711cd2..8a7f86c2139a 100644
> --- a/fs/fhandle.c
> +++ b/fs/fhandle.c
> @@ -202,7 +202,7 @@ static int handle_to_path(int mountdirfd, struct file_handle __user *ufh,
>  	*handle = f_handle;
>  	if (copy_from_user(&handle->f_handle,
>  			   &ufh->f_handle,
> -			   struct_size(ufh, f_handle, f_handle.handle_bytes))) {
> +			   f_handle.handle_bytes)) {
>  		retval = -EFAULT;
>  		goto out_handle;
>  	}

cc'ing Gustavo, since it looks like his patch in -next is what broke
this.

-- 
Jeff Layton <jlayton@kernel.org>