| Message ID | 20190311184105.40101-1-cpaasch@apple.com |
|---|---|
| State | Accepted |
| Delegated to: | David Miller |
| Headers | show |
| Series | tcp: Don't access TCP_SKB_CB before initializing it | expand |
On 03/11/2019 11:41 AM, Christoph Paasch wrote: > Since commit eeea10b83a13 ("tcp: add > tcp_v4_fill_cb()/tcp_v4_restore_cb()"), tcp_vX_fill_cb is only called > after tcp_filter(). That means, TCP_SKB_CB(skb)->end_seq still points to > the IP-part of the cb. > > We thus should not mock with it, as this can trigger bugs (thanks > syzkaller): > [ 12.349396] ================================================================== > [ 12.350188] BUG: KASAN: slab-out-of-bounds in ip6_datagram_recv_specific_ctl+0x19b3/0x1a20 > [ 12.351035] Read of size 1 at addr ffff88006adbc208 by task test_ip6_datagr/1799 > > Setting end_seq is actually no more necessary in tcp_filter as it gets > initialized later on in tcp_vX_fill_cb. > > Cc: Eric Dumazet <edumazet@google.com> > Fixes: eeea10b83a13 ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()") > Signed-off-by: Christoph Paasch <cpaasch@apple.com> > --- > net/ipv4/tcp_ipv4.c | 9 +-------- > 1 file changed, 1 insertion(+), 8 deletions(-) Good catch, thanks ! Signed-off-by: Eric Dumazet <edumazet@google.com>
From: Christoph Paasch <cpaasch@apple.com> Date: Mon, 11 Mar 2019 11:41:05 -0700 > Since commit eeea10b83a13 ("tcp: add > tcp_v4_fill_cb()/tcp_v4_restore_cb()"), tcp_vX_fill_cb is only called > after tcp_filter(). That means, TCP_SKB_CB(skb)->end_seq still points to > the IP-part of the cb. > > We thus should not mock with it, as this can trigger bugs (thanks > syzkaller): > [ 12.349396] ================================================================== > [ 12.350188] BUG: KASAN: slab-out-of-bounds in ip6_datagram_recv_specific_ctl+0x19b3/0x1a20 > [ 12.351035] Read of size 1 at addr ffff88006adbc208 by task test_ip6_datagr/1799 > > Setting end_seq is actually no more necessary in tcp_filter as it gets > initialized later on in tcp_vX_fill_cb. > > Cc: Eric Dumazet <edumazet@google.com> > Fixes: eeea10b83a13 ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()") > Signed-off-by: Christoph Paasch <cpaasch@apple.com> Applied.
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 831d844a27ca..277d71239d75 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1734,15 +1734,8 @@ EXPORT_SYMBOL(tcp_add_backlog); int tcp_filter(struct sock *sk, struct sk_buff *skb) { struct tcphdr *th = (struct tcphdr *)skb->data; - unsigned int eaten = skb->len; - int err; - err = sk_filter_trim_cap(sk, skb, th->doff * 4); - if (!err) { - eaten -= skb->len; - TCP_SKB_CB(skb)->end_seq -= eaten; - } - return err; + return sk_filter_trim_cap(sk, skb, th->doff * 4); } EXPORT_SYMBOL(tcp_filter);
Since commit eeea10b83a13 ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()"), tcp_vX_fill_cb is only called after tcp_filter(). That means, TCP_SKB_CB(skb)->end_seq still points to the IP-part of the cb. We thus should not mock with it, as this can trigger bugs (thanks syzkaller): [ 12.349396] ================================================================== [ 12.350188] BUG: KASAN: slab-out-of-bounds in ip6_datagram_recv_specific_ctl+0x19b3/0x1a20 [ 12.351035] Read of size 1 at addr ffff88006adbc208 by task test_ip6_datagr/1799 Setting end_seq is actually no more necessary in tcp_filter as it gets initialized later on in tcp_vX_fill_cb. Cc: Eric Dumazet <edumazet@google.com> Fixes: eeea10b83a13 ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()") Signed-off-by: Christoph Paasch <cpaasch@apple.com> --- net/ipv4/tcp_ipv4.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-)