Re: [PATCH] crypto_user: Fix out-of-bounds read

[Dan Carpenter]

On Tue, Apr 22, 2014 at 12:30:28PM -0700, Andy Lutomirski wrote:
> This is unlikely to be exploitable for anything except an OOPS.
> 
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Andy Lutomirski <luto@xxxxxxxxxxxxxx>
> ---
> 
> Notes:
>     This is entirely untested, but it looks obviously correct to me.
> 
>  crypto/crypto_user.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/crypto/crypto_user.c b/crypto/crypto_user.c
> index 1512e41..bc7c4b5 100644
> --- a/crypto/crypto_user.c
> +++ b/crypto/crypto_user.c
> @@ -460,7 +460,7 @@ static int crypto_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
>  	int type, err;
>  
>  	type = nlh->nlmsg_type;
> -	if (type > CRYPTO_MSG_MAX)
> +	if (type < CRYPTO_MSG_BASE || type > CRYPTO_MSG_MAX)
>  		return -EINVAL;

Adding a check here is obviously harmless but I think this is only
called from netlink_rcv_skb() which already checks:

	if (nlh->nlmsg_type < NLMSG_MIN_TYPE)
		goto ack;

NLMSG_MIN_TYPE is 0x10 as well, so I don't think we can hit your
condition.

Your patch freaked me out a little because this is one of the bugs that
I should have caught throught static analysis.

regards,
dan carpenter

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/