On 12/01/2017 11:48 AM, Michal Kubecek wrote:

On Thu, Nov 30, 2017 at 10:37:40AM -0800, Guenter Roeck wrote:

...

Hi,

The fix for CVE-2017-16939 has been applied to v4.9.y, but not to v4.4.y and older kernels. However, I confirmed that running the published POC (see https://blogs.securiteam.com/index.php/archives/3535) does crash a 4.4 kernel.

I confirmed that the following two patches fix the problem in v4.4.y. Please consider applying them to v4.4.y (and possibly v3.18.y).

fc9e50f5a5a4e ("netlink: add a start callback for starting a netlink dump") 1137b5e2529a8 ("ipsec: Fix aborted xfrm policy dump crash")

My apologies for the noise if this is already under consideration.

It's a bit too big hammer. As Nicolai Stange noticed when we were

The hammer is just as big as the upstream hammer. Personally I prefer the upstream patch; I don't see a reason to deviate from upstream just because the upstream solution is more complex than necessary.

handling this for SLE12 (where fc9e50f5a5a4e would break kABI), it's

I didn't know that this is even a concern for stable releases. Is there some guideline that kABI changes should be avoided in stable releases ?

Thanks, Guenter

much simpler to use the flag we already have in cb->args[0] to let xfrm_dump_policy_done() call xfrm_policy_walk_done() only if the walk structure has been initialized. Thus all you need is the patch below.

Michal Kubecek

diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 7a5a64e70b4d..c01c7a7eb4d3 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -1655,7 +1655,9 @@ static int xfrm_dump_policy_done(struct netlink_callback *cb) struct xfrm_policy_walk *walk = (struct xfrm_policy_walk *) &cb->args[1]; struct net *net = sock_net(cb->skb->sk);

• xfrm_policy_walk_done(walk, net);

• /* cb->args[0] is set when walk is initialized */
• if (cb->args[0])

• xfrm_policy_walk_done(walk, net);
  

  return 0; }