Re: [PATCH net] ipv4: icmp: reject non-unicast input routes in icmp_route_lookup
"dongchenchen (A)" <[email protected]>
| Newsgroups | org.kernel.vger.netdev |
|---|---|
| Message-ID | <[email protected]> |
> When the forward output route cannot be used in icmp_route_lookup(),
> it enters the "reverse path" and calls ip_route_input() on fl4_dec.daddr,
> the original packet's source address.
>
> ip_route_input() only returns an error for truly invalid packets. For
> local, broadcast, multicast and unreachable addresses it still succeeds
> and returns an input route whose dst.output is set to ip_rt_bug(). The
> existing check only rejects RTN_LOCAL routes, so the other route types
> can still be returned and later used for output, syzkaller triggering a
> WARN_ON_ONCE() in ip_rt_bug() as bellow:
>
> ------------[ cut here ]------------
> WARNING: net/ipv4/route.c:1273 at ip_rt_bug+0x14/0x20
> RIP: 0010:ip_rt_bug+0x14/0x20
> Call Trace:
> ip_push_pending_frames+0xfa/0x100
> __icmp_send+0x905/0xf10
> ip_options_compile+0xc0/0xd0
> ip_rcv_finish_core+0x321/0xae0
> ip_rcv+0x1de/0x260
> __netif_receive_skb_one_core+0x11a/0x130
> netif_receive_skb+0x7b/0x260
> tun_get_user+0x11bf/0x1c10
> ------------[ cut here ]------------
>
> If the original source address is unroutable, ip_route_input() returns a
> RTN_UNREACHABLE input route with err == 0, and the ICMP reply is then sent
> through ip_rt_bug().
>
> Only RTN_UNICAST input routes have dst.output set to ip_output() and are
> safe to use for output. Reject any input route that is not RTN_UNICAST.
>
> Fixes: 8b7817f3a959 ("[IPSEC]: Add ICMP host relookup support")
> Signed-off-by: Dong Chenchen <[email protected]>
> ---
> net/ipv4/icmp.c | 12 +++++++-----
> 1 file changed, 7 insertions(+), 5 deletions(-)
>
> diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
> index 0caedfc7ca92..d58ac2f23f9f 100644
> --- a/net/ipv4/icmp.c
> +++ b/net/ipv4/icmp.c
> @@ -584,12 +584,14 @@ static struct rtable *icmp_route_lookup(struct net *net, struct flowi4 *fl4,
> * At this point, fl4_dec.daddr should NOT be local (we
> * checked fl4_dec.saddr above). However, a race condition
> * may occur if the address is added to the interface
> - * concurrently. In that case, ip_route_input() returns a
> - * LOCAL route with dst.output=ip_rt_bug, which must not
> - * be used for output.
> + * concurrently, or if fl4_dec.daddr is otherwise unroutable.
> + * In that case, ip_route_input() returns an input route
> + * (RTN_LOCAL, RTN_BROADCAST, RTN_UNREACHABLE or
> + * RTN_MULTICAST) with dst.output set to ip_rt_bug(), which
> + * must not be used for output.
> */
> - if (!err && rt2 && rt2->rt_type == RTN_LOCAL) {
> - net_warn_ratelimited("detected local route for %pI4 during ICMP sending, src %pI4\n",
> + if (!err && rt2 && rt2->rt_type != RTN_UNICAST) {
sorry, this judgement will reject bforward packet.
broadcast and multicast rt with ip_rt_bug has been reject after
icmp_route_lookup.We only need to add a judgment for RTN_UNREACHABLE
v2 will be sent.
best regards
Dong Chenchen> + net_warn_ratelimited("detected unusable input route
for %pI4 during ICMP sending, src %pI4\n",
> &fl4_dec.daddr, &fl4_dec.saddr);
> dst_release(&rt2->dst);
> err = -EINVAL;