From d1a6cba576fc7c43e476538fe5aa72fe04bd80e1 Mon Sep 17 00:00:00 2001 From: Liping Zhang Date: Tue, 6 Sep 2016 22:31:02 +0800 Subject: netfilter: nft_chain_route: re-route before skb is queued to userspace Imagine such situation, user add the following nft rules, and queue the packets to userspace for further check: # ip rule add fwmark 0x0/0x1 lookup eth0 # ip rule add fwmark 0x1/0x1 lookup eth1 # nft add table filter # nft add chain filter output {type route hook output priority 0 \;} # nft add rule filter output mark set 0x1 # nft add rule filter output queue num 0 But after we reinject the skbuff, the packet will be sent via the wrong route, i.e. in this case, the packet will be routed via eth0 table, not eth1 table. Because we skip to do re-route when verdict is NF_QUEUE, even if the mark was changed. Acctually, we should not touch sk_buff if verdict is NF_DROP or NF_STOLEN, and when re-route fails, return NF_DROP with error code. This is consistent with the mangle table in iptables. Signed-off-by: Liping Zhang Signed-off-by: Pablo Neira Ayuso --- net/ipv4/netfilter/nft_chain_route_ipv4.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'net/ipv4') diff --git a/net/ipv4/netfilter/nft_chain_route_ipv4.c b/net/ipv4/netfilter/nft_chain_route_ipv4.c index 2375b0a..30493be 100644 --- a/net/ipv4/netfilter/nft_chain_route_ipv4.c +++ b/net/ipv4/netfilter/nft_chain_route_ipv4.c @@ -31,6 +31,7 @@ static unsigned int nf_route_table_hook(void *priv, __be32 saddr, daddr; u_int8_t tos; const struct iphdr *iph; + int err; /* root is playing with raw sockets. */ if (skb->len < sizeof(struct iphdr) || @@ -46,15 +47,17 @@ static unsigned int nf_route_table_hook(void *priv, tos = iph->tos; ret = nft_do_chain(&pkt, priv); - if (ret != NF_DROP && ret != NF_QUEUE) { + if (ret != NF_DROP && ret != NF_STOLEN) { iph = ip_hdr(skb); if (iph->saddr != saddr || iph->daddr != daddr || skb->mark != mark || - iph->tos != tos) - if (ip_route_me_harder(state->net, skb, RTN_UNSPEC)) - ret = NF_DROP; + iph->tos != tos) { + err = ip_route_me_harder(state->net, skb, RTN_UNSPEC); + if (err < 0) + ret = NF_DROP_ERR(err); + } } return ret; } -- cgit v1.1 From 1fb81e09d487656aa23f2acb1232c7f56b4c2367 Mon Sep 17 00:00:00 2001 From: "thomas.zeitlhofer+lkml@ze-it.at" Date: Wed, 7 Sep 2016 20:40:38 +0200 Subject: vti: use right inner_mode for inbound inter address family policy checks In case of inter address family tunneling (IPv6 over vti4 or IPv4 over vti6), the inbound policy checks in vti_rcv_cb() and vti6_rcv_cb() are using the wrong address family. As a result, all inbound inter address family traffic is dropped. Use the xfrm_ip2inner_mode() helper, as done in xfrm_input() (i.e., also increment LINUX_MIB_XFRMINSTATEMODEERROR in case of error), to select the inner_mode that contains the right address family for the inbound policy checks. Signed-off-by: Thomas Zeitlhofer Signed-off-by: Steffen Klassert --- net/ipv4/ip_vti.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) (limited to 'net/ipv4') diff --git a/net/ipv4/ip_vti.c b/net/ipv4/ip_vti.c index cc701fa..5d7944f 100644 --- a/net/ipv4/ip_vti.c +++ b/net/ipv4/ip_vti.c @@ -88,6 +88,7 @@ static int vti_rcv_cb(struct sk_buff *skb, int err) struct net_device *dev; struct pcpu_sw_netstats *tstats; struct xfrm_state *x; + struct xfrm_mode *inner_mode; struct ip_tunnel *tunnel = XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4; u32 orig_mark = skb->mark; int ret; @@ -105,7 +106,19 @@ static int vti_rcv_cb(struct sk_buff *skb, int err) } x = xfrm_input_state(skb); - family = x->inner_mode->afinfo->family; + + inner_mode = x->inner_mode; + + if (x->sel.family == AF_UNSPEC) { + inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol); + if (inner_mode == NULL) { + XFRM_INC_STATS(dev_net(skb->dev), + LINUX_MIB_XFRMINSTATEMODEERROR); + return -EINVAL; + } + } + + family = inner_mode->afinfo->family; skb->mark = be32_to_cpu(tunnel->parms.i_key); ret = xfrm_policy_check(NULL, XFRM_POLICY_IN, skb, family); -- cgit v1.1 From d6f64d725bac20df66b2eacd847fc41d7a1905e0 Mon Sep 17 00:00:00 2001 From: Mark Tomlinson Date: Thu, 15 Sep 2016 11:40:05 +1200 Subject: net: VRF: Pass original iif to ip_route_input() The function ip_rcv_finish() calls l3mdev_ip_rcv(). On any VRF except the global VRF, this replaces skb->dev with the VRF master interface. When calling ip_route_input_noref() from here, the checks for forwarding look at this master device instead of the initial ingress interface. This will allow packets to be routed which normally would be dropped. For example, an interface that is not assigned an IP address should drop packets, but because the checking is against the master device, the packet will be forwarded. The fix here is to still call l3mdev_ip_rcv(), but remember the initial net_device. This is passed to the other functions within ip_rcv_finish, so they still see the original interface. Signed-off-by: Mark Tomlinson Acked-by: David Ahern Signed-off-by: David S. Miller --- net/ipv4/ip_input.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'net/ipv4') diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c index 4b351af..d6feabb 100644 --- a/net/ipv4/ip_input.c +++ b/net/ipv4/ip_input.c @@ -312,6 +312,7 @@ static int ip_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb) { const struct iphdr *iph = ip_hdr(skb); struct rtable *rt; + struct net_device *dev = skb->dev; /* if ingress device is enslaved to an L3 master device pass the * skb to its handler for processing @@ -341,7 +342,7 @@ static int ip_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb) */ if (!skb_valid_dst(skb)) { int err = ip_route_input_noref(skb, iph->daddr, iph->saddr, - iph->tos, skb->dev); + iph->tos, dev); if (unlikely(err)) { if (err == -EXDEV) __NET_INC_STATS(net, LINUX_MIB_IPRPFILTER); @@ -370,7 +371,7 @@ static int ip_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb) __IP_UPD_PO_STATS(net, IPSTATS_MIB_INBCAST, skb->len); } else if (skb->pkt_type == PACKET_BROADCAST || skb->pkt_type == PACKET_MULTICAST) { - struct in_device *in_dev = __in_dev_get_rcu(skb->dev); + struct in_device *in_dev = __in_dev_get_rcu(dev); /* RFC 1122 3.3.6: * -- cgit v1.1 From ffb4d6c8508657824bcef68a36b2a0f9d8c09d10 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 15 Sep 2016 08:12:33 -0700 Subject: tcp: fix overflow in __tcp_retransmit_skb() If a TCP socket gets a large write queue, an overflow can happen in a test in __tcp_retransmit_skb() preventing all retransmits. The flow then stalls and resets after timeouts. Tested: sysctl -w net.core.wmem_max=1000000000 netperf -H dest -- -s 1000000000 Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_output.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/ipv4') diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index bdaef7f..f53d0cc 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2605,7 +2605,8 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) * copying overhead: fragmentation, tunneling, mangling etc. */ if (atomic_read(&sk->sk_wmem_alloc) > - min(sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), sk->sk_sndbuf)) + min_t(u32, sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), + sk->sk_sndbuf)) return -EAGAIN; if (skb_still_in_host_queue(sk, skb)) -- cgit v1.1 From b5036cd4ed3173ab8cdbc85e2ba74acf46bafb51 Mon Sep 17 00:00:00 2001 From: Nikolay Aleksandrov Date: Tue, 20 Sep 2016 16:17:22 +0200 Subject: ipmr, ip6mr: return lastuse relative to now When I introduced the lastuse member I made a subtle error because it was returned as an absolute value but that is meaningless to user-space as it doesn't allow to see how old exactly an entry is. Let's make it similar to how the bridge returns such values and make it relative to "now" (jiffies). This allows us to show the actual age of the entries and is much more useful (e.g. user-space daemons can age out entries, iproute2 can display the lastuse properly). Fixes: 43b9e1274060 ("net: ipmr/ip6mr: add support for keeping an entry age") Reported-by: Satish Ashok Signed-off-by: Nikolay Aleksandrov Signed-off-by: David S. Miller --- net/ipv4/ipmr.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'net/ipv4') diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index 2625332..a87bcd2 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -2076,6 +2076,7 @@ static int __ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb, struct rta_mfc_stats mfcs; struct nlattr *mp_attr; struct rtnexthop *nhp; + unsigned long lastuse; int ct; /* If cache is unresolved, don't try to parse IIF and OIF */ @@ -2105,12 +2106,14 @@ static int __ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb, nla_nest_end(skb, mp_attr); + lastuse = READ_ONCE(c->mfc_un.res.lastuse); + lastuse = time_after_eq(jiffies, lastuse) ? jiffies - lastuse : 0; + mfcs.mfcs_packets = c->mfc_un.res.pkt; mfcs.mfcs_bytes = c->mfc_un.res.bytes; mfcs.mfcs_wrong_if = c->mfc_un.res.wrong_if; if (nla_put_64bit(skb, RTA_MFC_STATS, sizeof(mfcs), &mfcs, RTA_PAD) || - nla_put_u64_64bit(skb, RTA_EXPIRES, - jiffies_to_clock_t(c->mfc_un.res.lastuse), + nla_put_u64_64bit(skb, RTA_EXPIRES, jiffies_to_clock_t(lastuse), RTA_PAD)) return -EMSGSIZE; -- cgit v1.1 From adb03115f4590baa280ddc440a8eff08a6be0cb7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 20 Sep 2016 18:06:17 -0700 Subject: net: get rid of an signed integer overflow in ip_idents_reserve() Jiri Pirko reported an UBSAN warning happening in ip_idents_reserve() [] UBSAN: Undefined behaviour in ./arch/x86/include/asm/atomic.h:156:11 [] signed integer overflow: [] -2117905507 + -695755206 cannot be represented in type 'int' Since we do not have uatomic_add_return() yet, use atomic_cmpxchg() so that the arithmetics can be done using unsigned int. Fixes: 04ca6973f7c1 ("ip: make IP identifiers less predictable") Signed-off-by: Eric Dumazet Reported-by: Jiri Pirko Signed-off-by: David S. Miller --- net/ipv4/route.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'net/ipv4') diff --git a/net/ipv4/route.c b/net/ipv4/route.c index a1f2830..b5b47a2 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -476,12 +476,18 @@ u32 ip_idents_reserve(u32 hash, int segs) atomic_t *p_id = ip_idents + hash % IP_IDENTS_SZ; u32 old = ACCESS_ONCE(*p_tstamp); u32 now = (u32)jiffies; - u32 delta = 0; + u32 new, delta = 0; if (old != now && cmpxchg(p_tstamp, old, now) == old) delta = prandom_u32_max(now - old); - return atomic_add_return(segs + delta, p_id) - segs; + /* Do not use atomic_add_return() as it makes UBSAN unhappy */ + do { + old = (u32)atomic_read(p_id); + new = old + delta + segs; + } while (atomic_cmpxchg(p_id, old, new) != old); + + return new - segs; } EXPORT_SYMBOL(ip_idents_reserve); -- cgit v1.1 From de1d657816c6fbb70f07b01d50ec669dff0d4e60 Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Wed, 21 Sep 2016 16:16:14 -0700 Subject: tcp: fix under-accounting retransmit SNMP counters This patch fixes these under-accounting SNMP rtx stats LINUX_MIB_TCPFORWARDRETRANS LINUX_MIB_TCPFASTRETRANS LINUX_MIB_TCPSLOWSTARTRETRANS when retransmitting TSO packets Fixes: 10d3be569243 ("tcp-tso: do not split TSO packets at retransmit time") Signed-off-by: Yuchung Cheng Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_output.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/ipv4') diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index f53d0cc..e15ec82 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2831,7 +2831,7 @@ begin_fwd: if (tcp_retransmit_skb(sk, skb, segs)) return; - NET_INC_STATS(sock_net(sk), mib_idx); + NET_ADD_STATS(sock_net(sk), mib_idx, tcp_skb_pcount(skb)); if (tcp_in_cwnd_reduction(sk)) tp->prr_out += tcp_skb_pcount(skb); -- cgit v1.1 From 7e32b44361abc77fbc01f2b97b045c405b2583e5 Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Wed, 21 Sep 2016 16:16:15 -0700 Subject: tcp: properly account Fast Open SYN-ACK retrans Since the TFO socket is accepted right off SYN-data, the socket owner can call getsockopt(TCP_INFO) to collect ongoing SYN-ACK retransmission or timeout stats (i.e., tcpi_total_retrans, tcpi_retransmits). Currently those stats are only updated upon handshake completes. This patch fixes it. Signed-off-by: Yuchung Cheng Signed-off-by: Eric Dumazet Signed-off-by: Neal Cardwell Signed-off-by: Soheil Hassas Yeganeh Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 2 +- net/ipv4/tcp_output.c | 2 ++ net/ipv4/tcp_timer.c | 1 + 3 files changed, 4 insertions(+), 1 deletion(-) (limited to 'net/ipv4') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 3ebf45b..08323bd 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5885,7 +5885,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) * so release it. */ if (req) { - tp->total_retrans = req->num_retrans; + inet_csk(sk)->icsk_retransmits = 0; reqsk_fastopen_remove(sk, req, false); } else { /* Make sure socket is routed, for correct metrics. */ diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index e15ec82..5288cec 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -3568,6 +3568,8 @@ int tcp_rtx_synack(const struct sock *sk, struct request_sock *req) if (!res) { __TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS); __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS); + if (unlikely(tcp_passive_fastopen(sk))) + tcp_sk(sk)->total_retrans++; } return res; } diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index d84930b..f712b41 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -384,6 +384,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk) */ inet_rtx_syn_ack(sk, req); req->num_timeout++; + icsk->icsk_retransmits++; inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, TCP_TIMEOUT_INIT << req->num_timeout, TCP_RTO_MAX); } -- cgit v1.1