Re: possible stack corruption in icmp_send (__stack_chk_fail)

On Wed, Feb 17, 2021 at 5:56 PM Jason A. Donenfeld [off-list ref] wrote:

Hi Willem,

On Wed, Feb 17, 2021 at 11:27 PM Willem de Bruijn
[off-list ref] wrote:

quoted

A vmlinux image might help. I couldn't find one for this kernel.

https://data.zx2c4.com/icmp_send-crash-e03b4a42-706a-43bf-bc40-1f15966b3216.tar.xz
has .debs with vmlinuz in there, which you can extract to vmlinux, as
well as my own vmlinux elf construction with the symbols added back in
by extracting them from kallsyms. That's the best I've been able to
do, as all of this is coming from somebody random emailing me.

quoted

But could it be
that the forwarded packet is not sensible IPv4? The skb->protocol is
inferred in wg_packet_consume_data_done->ip_tunnel_parse_protocol.

The wg calls to icmp_ndo_send are gated by checking skb->protocol:

        if (skb->protocol == htons(ETH_P_IP))
               icmp_ndo_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_UNREACH, 0);
       else if (skb->protocol == htons(ETH_P_IPV6))
               icmpv6_ndo_send(skb, ICMPV6_DEST_UNREACH,
ICMPV6_ADDR_UNREACH, 0);

On the other hand, that code is hit on an error path when
wg_check_packet_protocol returns false:

static inline bool wg_check_packet_protocol(struct sk_buff *skb)
{
       __be16 real_protocol = ip_tunnel_parse_protocol(skb);
       return real_protocol && skb->protocol == real_protocol;
}

So that means, at least in theory, icmp_ndo_send could be called with
skb->protocol != ip_tunnel_parse_protocol(skb). I guess I can address
that. But... is it actually a problem?

For this forwarded packet that arrived on a wireguard tunnel,
skb->protocol was originally also set by ip_tunnel_parse_protocol.
So likely not.

The other issue seems more like a real bug. wg_xmit calling
icmp_ndo_send without clearing IPCB first.