Re: [PATCH] net: recvmsg: Unconditionally zero struct sockaddr_storage

[Ben Hutchings <ben@decadent.org.uk>]

[-- Attachment #1: Type: text/plain, Size: 1612 bytes --]

On Tue, 2017-10-31 at 09:14 -0700, Kees Cook wrote:
> Some protocols do not correctly wipe the contents of the on-stack
> struct sockaddr_storage sent down into recvmsg() (e.g. SCTP), and leak
> kernel stack contents to userspace. This wipes it unconditionally before
> per-protocol handlers run.
> 
> Note that leaks like this are mitigated by building with
> CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL=y
> 
> Reported-by: Alexander Potapenko <glider@google.com>
> Cc: "David S. Miller" <davem@davemloft.net>
> Cc: netdev@vger.kernel.org
> Signed-off-by: Kees Cook <keescook@chromium.org>
> ---
>  net/socket.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/net/socket.c b/net/socket.c
> index c729625eb5d3..34183f4fbdf8 100644
> --- a/net/socket.c
> +++ b/net/socket.c
> @@ -2188,6 +2188,7 @@ static int ___sys_recvmsg(struct socket *sock, struct user_msghdr __user *msg,
>  	struct sockaddr __user *uaddr;
>  	int __user *uaddr_len = COMPAT_NAMELEN(msg);
>  
> +	memset(&addr, 0, sizeof(addr));

That's a fairly large structure (128 bytes), most of which won't
normally be used.  We already initialise msg_namelen to 0 before
calling the per-protocol handler, which means by default nothing leaks.
 Only cases where msg_namelen is set but msg_name[] is not initialised
up to that length are a problem.  I would have thought they were not
too hard to find and fix.

Ben.

>  	msg_sys->msg_name = &addr;
>  
>  	if (MSG_CMSG_COMPAT & flags)
> -- 
> 2.7.4
> 
> 
-- 
Ben Hutchings
It is a miracle that curiosity survives formal education. - Albert
Einstein


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]