aboutsummaryrefslogtreecommitdiffstats
path: root/net
diff options
context:
space:
mode:
Diffstat (limited to 'net')
-rw-r--r--net/bridge/netfilter/nft_reject_bridge.c29
-rw-r--r--net/caif/chnl_net.c1
-rw-r--r--net/core/dev.c37
-rw-r--r--net/core/rtnetlink.c6
-rw-r--r--net/ipv4/ip_output.c29
-rw-r--r--net/ipv4/route.c3
-rw-r--r--net/ipv4/tcp_bic.c2
-rw-r--r--net/ipv4/tcp_cong.c32
-rw-r--r--net/ipv4/tcp_cubic.c39
-rw-r--r--net/ipv4/tcp_ipv4.c37
-rw-r--r--net/ipv4/tcp_scalable.c3
-rw-r--r--net/ipv4/tcp_veno.c2
-rw-r--r--net/ipv4/tcp_yeah.c2
-rw-r--r--net/ipv6/ip6_gre.c4
-rw-r--r--net/ipv6/ip6_output.c14
-rw-r--r--net/ipv6/output_core.c41
-rw-r--r--net/ipv6/sit.c8
-rw-r--r--net/ipv6/udp_offload.c10
-rw-r--r--net/netfilter/ipvs/ip_vs_core.c33
-rw-r--r--net/netfilter/nf_tables_api.c28
-rw-r--r--net/netfilter/nft_masq.c26
-rw-r--r--net/netfilter/nft_nat.c40
-rw-r--r--net/netfilter/nft_redir.c25
-rw-r--r--net/netlink/af_netlink.c4
-rw-r--r--net/rds/sysctl.c4
-rw-r--r--net/sched/cls_api.c7
-rw-r--r--net/sched/sch_fq.c10
-rw-r--r--net/sctp/sm_make_chunk.c2
28 files changed, 274 insertions, 204 deletions
diff --git a/net/bridge/netfilter/nft_reject_bridge.c b/net/bridge/netfilter/nft_reject_bridge.c
index b0330aecbf97..3244aead0926 100644
--- a/net/bridge/netfilter/nft_reject_bridge.c
+++ b/net/bridge/netfilter/nft_reject_bridge.c
@@ -265,22 +265,12 @@ out:
265 data[NFT_REG_VERDICT].verdict = NF_DROP; 265 data[NFT_REG_VERDICT].verdict = NF_DROP;
266} 266}
267 267
268static int nft_reject_bridge_validate_hooks(const struct nft_chain *chain) 268static int nft_reject_bridge_validate(const struct nft_ctx *ctx,
269 const struct nft_expr *expr,
270 const struct nft_data **data)
269{ 271{
270 struct nft_base_chain *basechain; 272 return nft_chain_validate_hooks(ctx->chain, (1 << NF_BR_PRE_ROUTING) |
271 273 (1 << NF_BR_LOCAL_IN));
272 if (chain->flags & NFT_BASE_CHAIN) {
273 basechain = nft_base_chain(chain);
274
275 switch (basechain->ops[0].hooknum) {
276 case NF_BR_PRE_ROUTING:
277 case NF_BR_LOCAL_IN:
278 break;
279 default:
280 return -EOPNOTSUPP;
281 }
282 }
283 return 0;
284} 274}
285 275
286static int nft_reject_bridge_init(const struct nft_ctx *ctx, 276static int nft_reject_bridge_init(const struct nft_ctx *ctx,
@@ -290,7 +280,7 @@ static int nft_reject_bridge_init(const struct nft_ctx *ctx,
290 struct nft_reject *priv = nft_expr_priv(expr); 280 struct nft_reject *priv = nft_expr_priv(expr);
291 int icmp_code, err; 281 int icmp_code, err;
292 282
293 err = nft_reject_bridge_validate_hooks(ctx->chain); 283 err = nft_reject_bridge_validate(ctx, expr, NULL);
294 if (err < 0) 284 if (err < 0)
295 return err; 285 return err;
296 286
@@ -341,13 +331,6 @@ nla_put_failure:
341 return -1; 331 return -1;
342} 332}
343 333
344static int nft_reject_bridge_validate(const struct nft_ctx *ctx,
345 const struct nft_expr *expr,
346 const struct nft_data **data)
347{
348 return nft_reject_bridge_validate_hooks(ctx->chain);
349}
350
351static struct nft_expr_type nft_reject_bridge_type; 334static struct nft_expr_type nft_reject_bridge_type;
352static const struct nft_expr_ops nft_reject_bridge_ops = { 335static const struct nft_expr_ops nft_reject_bridge_ops = {
353 .type = &nft_reject_bridge_type, 336 .type = &nft_reject_bridge_type,
diff --git a/net/caif/chnl_net.c b/net/caif/chnl_net.c
index 4589ff67bfa9..67a4a36febd1 100644
--- a/net/caif/chnl_net.c
+++ b/net/caif/chnl_net.c
@@ -470,7 +470,6 @@ static int ipcaif_newlink(struct net *src_net, struct net_device *dev,
470 ASSERT_RTNL(); 470 ASSERT_RTNL();
471 caifdev = netdev_priv(dev); 471 caifdev = netdev_priv(dev);
472 caif_netlink_parms(data, &caifdev->conn_req); 472 caif_netlink_parms(data, &caifdev->conn_req);
473 dev_net_set(caifdev->netdev, src_net);
474 473
475 ret = register_netdevice(dev); 474 ret = register_netdevice(dev);
476 if (ret) 475 if (ret)
diff --git a/net/core/dev.c b/net/core/dev.c
index 171420e75b03..7fe82929f509 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2352,7 +2352,6 @@ EXPORT_SYMBOL(skb_checksum_help);
2352 2352
2353__be16 skb_network_protocol(struct sk_buff *skb, int *depth) 2353__be16 skb_network_protocol(struct sk_buff *skb, int *depth)
2354{ 2354{
2355 unsigned int vlan_depth = skb->mac_len;
2356 __be16 type = skb->protocol; 2355 __be16 type = skb->protocol;
2357 2356
2358 /* Tunnel gso handlers can set protocol to ethernet. */ 2357 /* Tunnel gso handlers can set protocol to ethernet. */
@@ -2366,35 +2365,7 @@ __be16 skb_network_protocol(struct sk_buff *skb, int *depth)
2366 type = eth->h_proto; 2365 type = eth->h_proto;
2367 } 2366 }
2368 2367
2369 /* if skb->protocol is 802.1Q/AD then the header should already be 2368 return __vlan_get_protocol(skb, type, depth);
2370 * present at mac_len - VLAN_HLEN (if mac_len > 0), or at
2371 * ETH_HLEN otherwise
2372 */
2373 if (type == htons(ETH_P_8021Q) || type == htons(ETH_P_8021AD)) {
2374 if (vlan_depth) {
2375 if (WARN_ON(vlan_depth < VLAN_HLEN))
2376 return 0;
2377 vlan_depth -= VLAN_HLEN;
2378 } else {
2379 vlan_depth = ETH_HLEN;
2380 }
2381 do {
2382 struct vlan_hdr *vh;
2383
2384 if (unlikely(!pskb_may_pull(skb,
2385 vlan_depth + VLAN_HLEN)))
2386 return 0;
2387
2388 vh = (struct vlan_hdr *)(skb->data + vlan_depth);
2389 type = vh->h_vlan_encapsulated_proto;
2390 vlan_depth += VLAN_HLEN;
2391 } while (type == htons(ETH_P_8021Q) ||
2392 type == htons(ETH_P_8021AD));
2393 }
2394
2395 *depth = vlan_depth;
2396
2397 return type;
2398} 2369}
2399 2370
2400/** 2371/**
@@ -5323,7 +5294,7 @@ void netdev_upper_dev_unlink(struct net_device *dev,
5323} 5294}
5324EXPORT_SYMBOL(netdev_upper_dev_unlink); 5295EXPORT_SYMBOL(netdev_upper_dev_unlink);
5325 5296
5326void netdev_adjacent_add_links(struct net_device *dev) 5297static void netdev_adjacent_add_links(struct net_device *dev)
5327{ 5298{
5328 struct netdev_adjacent *iter; 5299 struct netdev_adjacent *iter;
5329 5300
@@ -5348,7 +5319,7 @@ void netdev_adjacent_add_links(struct net_device *dev)
5348 } 5319 }
5349} 5320}
5350 5321
5351void netdev_adjacent_del_links(struct net_device *dev) 5322static void netdev_adjacent_del_links(struct net_device *dev)
5352{ 5323{
5353 struct netdev_adjacent *iter; 5324 struct netdev_adjacent *iter;
5354 5325
@@ -6656,7 +6627,7 @@ struct netdev_queue *dev_ingress_queue_create(struct net_device *dev)
6656 if (!queue) 6627 if (!queue)
6657 return NULL; 6628 return NULL;
6658 netdev_init_one_queue(dev, queue, NULL); 6629 netdev_init_one_queue(dev, queue, NULL);
6659 queue->qdisc = &noop_qdisc; 6630 RCU_INIT_POINTER(queue->qdisc, &noop_qdisc);
6660 queue->qdisc_sleeping = &noop_qdisc; 6631 queue->qdisc_sleeping = &noop_qdisc;
6661 rcu_assign_pointer(dev->ingress_queue, queue); 6632 rcu_assign_pointer(dev->ingress_queue, queue);
6662#endif 6633#endif
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 9cf6fe9ddc0c..446cbaf81185 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -2895,12 +2895,16 @@ static int rtnl_bridge_notify(struct net_device *dev, u16 flags)
2895 goto errout; 2895 goto errout;
2896 } 2896 }
2897 2897
2898 if (!skb->len)
2899 goto errout;
2900
2898 rtnl_notify(skb, net, 0, RTNLGRP_LINK, NULL, GFP_ATOMIC); 2901 rtnl_notify(skb, net, 0, RTNLGRP_LINK, NULL, GFP_ATOMIC);
2899 return 0; 2902 return 0;
2900errout: 2903errout:
2901 WARN_ON(err == -EMSGSIZE); 2904 WARN_ON(err == -EMSGSIZE);
2902 kfree_skb(skb); 2905 kfree_skb(skb);
2903 rtnl_set_sk_err(net, RTNLGRP_LINK, err); 2906 if (err)
2907 rtnl_set_sk_err(net, RTNLGRP_LINK, err);
2904 return err; 2908 return err;
2905} 2909}
2906 2910
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index b50861b22b6b..c373c0708d97 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1506,23 +1506,8 @@ static int ip_reply_glue_bits(void *dptr, char *to, int offset,
1506/* 1506/*
1507 * Generic function to send a packet as reply to another packet. 1507 * Generic function to send a packet as reply to another packet.
1508 * Used to send some TCP resets/acks so far. 1508 * Used to send some TCP resets/acks so far.
1509 *
1510 * Use a fake percpu inet socket to avoid false sharing and contention.
1511 */ 1509 */
1512static DEFINE_PER_CPU(struct inet_sock, unicast_sock) = { 1510void ip_send_unicast_reply(struct sock *sk, struct sk_buff *skb,
1513 .sk = {
1514 .__sk_common = {
1515 .skc_refcnt = ATOMIC_INIT(1),
1516 },
1517 .sk_wmem_alloc = ATOMIC_INIT(1),
1518 .sk_allocation = GFP_ATOMIC,
1519 .sk_flags = (1UL << SOCK_USE_WRITE_QUEUE),
1520 },
1521 .pmtudisc = IP_PMTUDISC_WANT,
1522 .uc_ttl = -1,
1523};
1524
1525void ip_send_unicast_reply(struct net *net, struct sk_buff *skb,
1526 const struct ip_options *sopt, 1511 const struct ip_options *sopt,
1527 __be32 daddr, __be32 saddr, 1512 __be32 daddr, __be32 saddr,
1528 const struct ip_reply_arg *arg, 1513 const struct ip_reply_arg *arg,
@@ -1532,9 +1517,8 @@ void ip_send_unicast_reply(struct net *net, struct sk_buff *skb,
1532 struct ipcm_cookie ipc; 1517 struct ipcm_cookie ipc;
1533 struct flowi4 fl4; 1518 struct flowi4 fl4;
1534 struct rtable *rt = skb_rtable(skb); 1519 struct rtable *rt = skb_rtable(skb);
1520 struct net *net = sock_net(sk);
1535 struct sk_buff *nskb; 1521 struct sk_buff *nskb;
1536 struct sock *sk;
1537 struct inet_sock *inet;
1538 int err; 1522 int err;
1539 1523
1540 if (__ip_options_echo(&replyopts.opt.opt, skb, sopt)) 1524 if (__ip_options_echo(&replyopts.opt.opt, skb, sopt))
@@ -1565,15 +1549,11 @@ void ip_send_unicast_reply(struct net *net, struct sk_buff *skb,
1565 if (IS_ERR(rt)) 1549 if (IS_ERR(rt))
1566 return; 1550 return;
1567 1551
1568 inet = &get_cpu_var(unicast_sock); 1552 inet_sk(sk)->tos = arg->tos;
1569 1553
1570 inet->tos = arg->tos;
1571 sk = &inet->sk;
1572 sk->sk_priority = skb->priority; 1554 sk->sk_priority = skb->priority;
1573 sk->sk_protocol = ip_hdr(skb)->protocol; 1555 sk->sk_protocol = ip_hdr(skb)->protocol;
1574 sk->sk_bound_dev_if = arg->bound_dev_if; 1556 sk->sk_bound_dev_if = arg->bound_dev_if;
1575 sock_net_set(sk, net);
1576 __skb_queue_head_init(&sk->sk_write_queue);
1577 sk->sk_sndbuf = sysctl_wmem_default; 1557 sk->sk_sndbuf = sysctl_wmem_default;
1578 err = ip_append_data(sk, &fl4, ip_reply_glue_bits, arg->iov->iov_base, 1558 err = ip_append_data(sk, &fl4, ip_reply_glue_bits, arg->iov->iov_base,
1579 len, 0, &ipc, &rt, MSG_DONTWAIT); 1559 len, 0, &ipc, &rt, MSG_DONTWAIT);
@@ -1589,13 +1569,10 @@ void ip_send_unicast_reply(struct net *net, struct sk_buff *skb,
1589 arg->csumoffset) = csum_fold(csum_add(nskb->csum, 1569 arg->csumoffset) = csum_fold(csum_add(nskb->csum,
1590 arg->csum)); 1570 arg->csum));
1591 nskb->ip_summed = CHECKSUM_NONE; 1571 nskb->ip_summed = CHECKSUM_NONE;
1592 skb_orphan(nskb);
1593 skb_set_queue_mapping(nskb, skb_get_queue_mapping(skb)); 1572 skb_set_queue_mapping(nskb, skb_get_queue_mapping(skb));
1594 ip_push_pending_frames(sk, &fl4); 1573 ip_push_pending_frames(sk, &fl4);
1595 } 1574 }
1596out: 1575out:
1597 put_cpu_var(unicast_sock);
1598
1599 ip_rt_put(rt); 1576 ip_rt_put(rt);
1600} 1577}
1601 1578
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index d58dd0ec3e53..52e1f2bf0ca2 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -966,6 +966,9 @@ static void __ip_rt_update_pmtu(struct rtable *rt, struct flowi4 *fl4, u32 mtu)
966 if (dst->dev->mtu < mtu) 966 if (dst->dev->mtu < mtu)
967 return; 967 return;
968 968
969 if (rt->rt_pmtu && rt->rt_pmtu < mtu)
970 return;
971
969 if (mtu < ip_rt_min_pmtu) 972 if (mtu < ip_rt_min_pmtu)
970 mtu = ip_rt_min_pmtu; 973 mtu = ip_rt_min_pmtu;
971 974
diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c
index bb395d46a389..c037644eafb7 100644
--- a/net/ipv4/tcp_bic.c
+++ b/net/ipv4/tcp_bic.c
@@ -150,7 +150,7 @@ static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
150 tcp_slow_start(tp, acked); 150 tcp_slow_start(tp, acked);
151 else { 151 else {
152 bictcp_update(ca, tp->snd_cwnd); 152 bictcp_update(ca, tp->snd_cwnd);
153 tcp_cong_avoid_ai(tp, ca->cnt); 153 tcp_cong_avoid_ai(tp, ca->cnt, 1);
154 } 154 }
155} 155}
156 156
diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
index 27ead0dd16bc..8670e68e2ce6 100644
--- a/net/ipv4/tcp_cong.c
+++ b/net/ipv4/tcp_cong.c
@@ -291,26 +291,32 @@ int tcp_set_congestion_control(struct sock *sk, const char *name)
291 * ABC caps N to 2. Slow start exits when cwnd grows over ssthresh and 291 * ABC caps N to 2. Slow start exits when cwnd grows over ssthresh and
292 * returns the leftover acks to adjust cwnd in congestion avoidance mode. 292 * returns the leftover acks to adjust cwnd in congestion avoidance mode.
293 */ 293 */
294void tcp_slow_start(struct tcp_sock *tp, u32 acked) 294u32 tcp_slow_start(struct tcp_sock *tp, u32 acked)
295{ 295{
296 u32 cwnd = tp->snd_cwnd + acked; 296 u32 cwnd = tp->snd_cwnd + acked;
297 297
298 if (cwnd > tp->snd_ssthresh) 298 if (cwnd > tp->snd_ssthresh)
299 cwnd = tp->snd_ssthresh + 1; 299 cwnd = tp->snd_ssthresh + 1;
300 acked -= cwnd - tp->snd_cwnd;
300 tp->snd_cwnd = min(cwnd, tp->snd_cwnd_clamp); 301 tp->snd_cwnd = min(cwnd, tp->snd_cwnd_clamp);
302
303 return acked;
301} 304}
302EXPORT_SYMBOL_GPL(tcp_slow_start); 305EXPORT_SYMBOL_GPL(tcp_slow_start);
303 306
304/* In theory this is tp->snd_cwnd += 1 / tp->snd_cwnd (or alternative w) */ 307/* In theory this is tp->snd_cwnd += 1 / tp->snd_cwnd (or alternative w),
305void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w) 308 * for every packet that was ACKed.
309 */
310void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w, u32 acked)
306{ 311{
312 tp->snd_cwnd_cnt += acked;
307 if (tp->snd_cwnd_cnt >= w) { 313 if (tp->snd_cwnd_cnt >= w) {
308 if (tp->snd_cwnd < tp->snd_cwnd_clamp) 314 u32 delta = tp->snd_cwnd_cnt / w;
309 tp->snd_cwnd++; 315
310 tp->snd_cwnd_cnt = 0; 316 tp->snd_cwnd_cnt -= delta * w;
311 } else { 317 tp->snd_cwnd += delta;
312 tp->snd_cwnd_cnt++;
313 } 318 }
319 tp->snd_cwnd = min(tp->snd_cwnd, tp->snd_cwnd_clamp);
314} 320}
315EXPORT_SYMBOL_GPL(tcp_cong_avoid_ai); 321EXPORT_SYMBOL_GPL(tcp_cong_avoid_ai);
316 322
@@ -329,11 +335,13 @@ void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
329 return; 335 return;
330 336
331 /* In "safe" area, increase. */ 337 /* In "safe" area, increase. */
332 if (tp->snd_cwnd <= tp->snd_ssthresh) 338 if (tp->snd_cwnd <= tp->snd_ssthresh) {
333 tcp_slow_start(tp, acked); 339 acked = tcp_slow_start(tp, acked);
340 if (!acked)
341 return;
342 }
334 /* In dangerous area, increase slowly. */ 343 /* In dangerous area, increase slowly. */
335 else 344 tcp_cong_avoid_ai(tp, tp->snd_cwnd, acked);
336 tcp_cong_avoid_ai(tp, tp->snd_cwnd);
337} 345}
338EXPORT_SYMBOL_GPL(tcp_reno_cong_avoid); 346EXPORT_SYMBOL_GPL(tcp_reno_cong_avoid);
339 347
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index 6b6002416a73..4b276d1ed980 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -93,9 +93,7 @@ struct bictcp {
93 u32 epoch_start; /* beginning of an epoch */ 93 u32 epoch_start; /* beginning of an epoch */
94 u32 ack_cnt; /* number of acks */ 94 u32 ack_cnt; /* number of acks */
95 u32 tcp_cwnd; /* estimated tcp cwnd */ 95 u32 tcp_cwnd; /* estimated tcp cwnd */
96#define ACK_RATIO_SHIFT 4 96 u16 unused;
97#define ACK_RATIO_LIMIT (32u << ACK_RATIO_SHIFT)
98 u16 delayed_ack; /* estimate the ratio of Packets/ACKs << 4 */
99 u8 sample_cnt; /* number of samples to decide curr_rtt */ 97 u8 sample_cnt; /* number of samples to decide curr_rtt */
100 u8 found; /* the exit point is found? */ 98 u8 found; /* the exit point is found? */
101 u32 round_start; /* beginning of each round */ 99 u32 round_start; /* beginning of each round */
@@ -114,7 +112,6 @@ static inline void bictcp_reset(struct bictcp *ca)
114 ca->bic_K = 0; 112 ca->bic_K = 0;
115 ca->delay_min = 0; 113 ca->delay_min = 0;
116 ca->epoch_start = 0; 114 ca->epoch_start = 0;
117 ca->delayed_ack = 2 << ACK_RATIO_SHIFT;
118 ca->ack_cnt = 0; 115 ca->ack_cnt = 0;
119 ca->tcp_cwnd = 0; 116 ca->tcp_cwnd = 0;
120 ca->found = 0; 117 ca->found = 0;
@@ -205,23 +202,30 @@ static u32 cubic_root(u64 a)
205/* 202/*
206 * Compute congestion window to use. 203 * Compute congestion window to use.
207 */ 204 */
208static inline void bictcp_update(struct bictcp *ca, u32 cwnd) 205static inline void bictcp_update(struct bictcp *ca, u32 cwnd, u32 acked)
209{ 206{
210 u32 delta, bic_target, max_cnt; 207 u32 delta, bic_target, max_cnt;
211 u64 offs, t; 208 u64 offs, t;
212 209
213 ca->ack_cnt++; /* count the number of ACKs */ 210 ca->ack_cnt += acked; /* count the number of ACKed packets */
214 211
215 if (ca->last_cwnd == cwnd && 212 if (ca->last_cwnd == cwnd &&
216 (s32)(tcp_time_stamp - ca->last_time) <= HZ / 32) 213 (s32)(tcp_time_stamp - ca->last_time) <= HZ / 32)
217 return; 214 return;
218 215
216 /* The CUBIC function can update ca->cnt at most once per jiffy.
217 * On all cwnd reduction events, ca->epoch_start is set to 0,
218 * which will force a recalculation of ca->cnt.
219 */
220 if (ca->epoch_start && tcp_time_stamp == ca->last_time)
221 goto tcp_friendliness;
222
219 ca->last_cwnd = cwnd; 223 ca->last_cwnd = cwnd;
220 ca->last_time = tcp_time_stamp; 224 ca->last_time = tcp_time_stamp;
221 225
222 if (ca->epoch_start == 0) { 226 if (ca->epoch_start == 0) {
223 ca->epoch_start = tcp_time_stamp; /* record beginning */ 227 ca->epoch_start = tcp_time_stamp; /* record beginning */
224 ca->ack_cnt = 1; /* start counting */ 228 ca->ack_cnt = acked; /* start counting */
225 ca->tcp_cwnd = cwnd; /* syn with cubic */ 229 ca->tcp_cwnd = cwnd; /* syn with cubic */
226 230
227 if (ca->last_max_cwnd <= cwnd) { 231 if (ca->last_max_cwnd <= cwnd) {
@@ -283,6 +287,7 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
283 if (ca->last_max_cwnd == 0 && ca->cnt > 20) 287 if (ca->last_max_cwnd == 0 && ca->cnt > 20)
284 ca->cnt = 20; /* increase cwnd 5% per RTT */ 288 ca->cnt = 20; /* increase cwnd 5% per RTT */
285 289
290tcp_friendliness:
286 /* TCP Friendly */ 291 /* TCP Friendly */
287 if (tcp_friendliness) { 292 if (tcp_friendliness) {
288 u32 scale = beta_scale; 293 u32 scale = beta_scale;
@@ -301,7 +306,6 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
301 } 306 }
302 } 307 }
303 308
304 ca->cnt = (ca->cnt << ACK_RATIO_SHIFT) / ca->delayed_ack;
305 if (ca->cnt == 0) /* cannot be zero */ 309 if (ca->cnt == 0) /* cannot be zero */
306 ca->cnt = 1; 310 ca->cnt = 1;
307} 311}
@@ -317,11 +321,12 @@ static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
317 if (tp->snd_cwnd <= tp->snd_ssthresh) { 321 if (tp->snd_cwnd <= tp->snd_ssthresh) {
318 if (hystart && after(ack, ca->end_seq)) 322 if (hystart && after(ack, ca->end_seq))
319 bictcp_hystart_reset(sk); 323 bictcp_hystart_reset(sk);
320 tcp_slow_start(tp, acked); 324 acked = tcp_slow_start(tp, acked);
321 } else { 325 if (!acked)
322 bictcp_update(ca, tp->snd_cwnd); 326 return;
323 tcp_cong_avoid_ai(tp, ca->cnt);
324 } 327 }
328 bictcp_update(ca, tp->snd_cwnd, acked);
329 tcp_cong_avoid_ai(tp, ca->cnt, acked);
325} 330}
326 331
327static u32 bictcp_recalc_ssthresh(struct sock *sk) 332static u32 bictcp_recalc_ssthresh(struct sock *sk)
@@ -411,20 +416,10 @@ static void hystart_update(struct sock *sk, u32 delay)
411 */ 416 */
412static void bictcp_acked(struct sock *sk, u32 cnt, s32 rtt_us) 417static void bictcp_acked(struct sock *sk, u32 cnt, s32 rtt_us)
413{ 418{
414 const struct inet_connection_sock *icsk = inet_csk(sk);
415 const struct tcp_sock *tp = tcp_sk(sk); 419 const struct tcp_sock *tp = tcp_sk(sk);
416 struct bictcp *ca = inet_csk_ca(sk); 420 struct bictcp *ca = inet_csk_ca(sk);
417 u32 delay; 421 u32 delay;
418 422
419 if (icsk->icsk_ca_state == TCP_CA_Open) {
420 u32 ratio = ca->delayed_ack;
421
422 ratio -= ca->delayed_ack >> ACK_RATIO_SHIFT;
423 ratio += cnt;
424
425 ca->delayed_ack = clamp(ratio, 1U, ACK_RATIO_LIMIT);
426 }
427
428 /* Some calls are for duplicates without timetamps */ 423 /* Some calls are for duplicates without timetamps */
429 if (rtt_us < 0) 424 if (rtt_us < 0)
430 return; 425 return;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index a3f72d7fc06c..d22f54482bab 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -683,7 +683,8 @@ static void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb)
683 arg.bound_dev_if = sk->sk_bound_dev_if; 683 arg.bound_dev_if = sk->sk_bound_dev_if;
684 684
685 arg.tos = ip_hdr(skb)->tos; 685 arg.tos = ip_hdr(skb)->tos;
686 ip_send_unicast_reply(net, skb, &TCP_SKB_CB(skb)->header.h4.opt, 686 ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
687 skb, &TCP_SKB_CB(skb)->header.h4.opt,
687 ip_hdr(skb)->saddr, ip_hdr(skb)->daddr, 688 ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
688 &arg, arg.iov[0].iov_len); 689 &arg, arg.iov[0].iov_len);
689 690
@@ -767,7 +768,8 @@ static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack,
767 if (oif) 768 if (oif)
768 arg.bound_dev_if = oif; 769 arg.bound_dev_if = oif;
769 arg.tos = tos; 770 arg.tos = tos;
770 ip_send_unicast_reply(net, skb, &TCP_SKB_CB(skb)->header.h4.opt, 771 ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk),
772 skb, &TCP_SKB_CB(skb)->header.h4.opt,
771 ip_hdr(skb)->saddr, ip_hdr(skb)->daddr, 773 ip_hdr(skb)->saddr, ip_hdr(skb)->daddr,
772 &arg, arg.iov[0].iov_len); 774 &arg, arg.iov[0].iov_len);
773 775
@@ -2428,14 +2430,39 @@ struct proto tcp_prot = {
2428}; 2430};
2429EXPORT_SYMBOL(tcp_prot); 2431EXPORT_SYMBOL(tcp_prot);
2430 2432
2433static void __net_exit tcp_sk_exit(struct net *net)
2434{
2435 int cpu;
2436
2437 for_each_possible_cpu(cpu)
2438 inet_ctl_sock_destroy(*per_cpu_ptr(net->ipv4.tcp_sk, cpu));
2439 free_percpu(net->ipv4.tcp_sk);
2440}
2441
2431static int __net_init tcp_sk_init(struct net *net) 2442static int __net_init tcp_sk_init(struct net *net)
2432{ 2443{
2444 int res, cpu;
2445
2446 net->ipv4.tcp_sk = alloc_percpu(struct sock *);
2447 if (!net->ipv4.tcp_sk)
2448 return -ENOMEM;
2449
2450 for_each_possible_cpu(cpu) {
2451 struct sock *sk;
2452
2453 res = inet_ctl_sock_create(&sk, PF_INET, SOCK_RAW,
2454 IPPROTO_TCP, net);
2455 if (res)
2456 goto fail;
2457 *per_cpu_ptr(net->ipv4.tcp_sk, cpu) = sk;
2458 }
2433 net->ipv4.sysctl_tcp_ecn = 2; 2459 net->ipv4.sysctl_tcp_ecn = 2;
2434 return 0; 2460 return 0;
2435}
2436 2461
2437static void __net_exit tcp_sk_exit(struct net *net) 2462fail:
2438{ 2463 tcp_sk_exit(net);
2464
2465 return res;
2439} 2466}
2440 2467
2441static void __net_exit tcp_sk_exit_batch(struct list_head *net_exit_list) 2468static void __net_exit tcp_sk_exit_batch(struct list_head *net_exit_list)
diff --git a/net/ipv4/tcp_scalable.c b/net/ipv4/tcp_scalable.c
index 6824afb65d93..333bcb2415ff 100644
--- a/net/ipv4/tcp_scalable.c
+++ b/net/ipv4/tcp_scalable.c
@@ -25,7 +25,8 @@ static void tcp_scalable_cong_avoid(struct sock *sk, u32 ack, u32 acked)
25 if (tp->snd_cwnd <= tp->snd_ssthresh) 25 if (tp->snd_cwnd <= tp->snd_ssthresh)
26 tcp_slow_start(tp, acked); 26 tcp_slow_start(tp, acked);
27 else 27 else
28 tcp_cong_avoid_ai(tp, min(tp->snd_cwnd, TCP_SCALABLE_AI_CNT)); 28 tcp_cong_avoid_ai(tp, min(tp->snd_cwnd, TCP_SCALABLE_AI_CNT),
29 1);
29} 30}
30 31
31static u32 tcp_scalable_ssthresh(struct sock *sk) 32static u32 tcp_scalable_ssthresh(struct sock *sk)
diff --git a/net/ipv4/tcp_veno.c b/net/ipv4/tcp_veno.c
index a4d2d2d88dca..112151eeee45 100644
--- a/net/ipv4/tcp_veno.c
+++ b/net/ipv4/tcp_veno.c
@@ -159,7 +159,7 @@ static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
159 /* In the "non-congestive state", increase cwnd 159 /* In the "non-congestive state", increase cwnd
160 * every rtt. 160 * every rtt.
161 */ 161 */
162 tcp_cong_avoid_ai(tp, tp->snd_cwnd); 162 tcp_cong_avoid_ai(tp, tp->snd_cwnd, 1);
163 } else { 163 } else {
164 /* In the "congestive state", increase cwnd 164 /* In the "congestive state", increase cwnd
165 * every other rtt. 165 * every other rtt.
diff --git a/net/ipv4/tcp_yeah.c b/net/ipv4/tcp_yeah.c
index cd7273218598..17d35662930d 100644
--- a/net/ipv4/tcp_yeah.c
+++ b/net/ipv4/tcp_yeah.c
@@ -92,7 +92,7 @@ static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
92 92
93 } else { 93 } else {
94 /* Reno */ 94 /* Reno */
95 tcp_cong_avoid_ai(tp, tp->snd_cwnd); 95 tcp_cong_avoid_ai(tp, tp->snd_cwnd, 1);
96 } 96 }
97 97
98 /* The key players are v_vegas.beg_snd_una and v_beg_snd_nxt. 98 /* The key players are v_vegas.beg_snd_una and v_beg_snd_nxt.
diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
index 13cda4c6313b..01ccc28a686f 100644
--- a/net/ipv6/ip6_gre.c
+++ b/net/ipv6/ip6_gre.c
@@ -417,7 +417,7 @@ static void ip6gre_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
417 if (code == ICMPV6_HDR_FIELD) 417 if (code == ICMPV6_HDR_FIELD)
418 teli = ip6_tnl_parse_tlv_enc_lim(skb, skb->data); 418 teli = ip6_tnl_parse_tlv_enc_lim(skb, skb->data);
419 419
420 if (teli && teli == info - 2) { 420 if (teli && teli == be32_to_cpu(info) - 2) {
421 tel = (struct ipv6_tlv_tnl_enc_lim *) &skb->data[teli]; 421 tel = (struct ipv6_tlv_tnl_enc_lim *) &skb->data[teli];
422 if (tel->encap_limit == 0) { 422 if (tel->encap_limit == 0) {
423 net_warn_ratelimited("%s: Too small encapsulation limit or routing loop in tunnel!\n", 423 net_warn_ratelimited("%s: Too small encapsulation limit or routing loop in tunnel!\n",
@@ -429,7 +429,7 @@ static void ip6gre_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
429 } 429 }
430 break; 430 break;
431 case ICMPV6_PKT_TOOBIG: 431 case ICMPV6_PKT_TOOBIG:
432 mtu = info - offset; 432 mtu = be32_to_cpu(info) - offset;
433 if (mtu < IPV6_MIN_MTU) 433 if (mtu < IPV6_MIN_MTU)
434 mtu = IPV6_MIN_MTU; 434 mtu = IPV6_MIN_MTU;
435 t->dev->mtu = mtu; 435 t->dev->mtu = mtu;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index ce69a12ae48c..d28f2a2efb32 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -537,20 +537,6 @@ static void ip6_copy_metadata(struct sk_buff *to, struct sk_buff *from)
537 skb_copy_secmark(to, from); 537 skb_copy_secmark(to, from);
538} 538}
539 539
540static void ipv6_select_ident(struct frag_hdr *fhdr, struct rt6_info *rt)
541{
542 static u32 ip6_idents_hashrnd __read_mostly;
543 u32 hash, id;
544
545 net_get_random_once(&ip6_idents_hashrnd, sizeof(ip6_idents_hashrnd));
546
547 hash = __ipv6_addr_jhash(&rt->rt6i_dst.addr, ip6_idents_hashrnd);
548 hash = __ipv6_addr_jhash(&rt->rt6i_src.addr, hash);
549
550 id = ip_idents_reserve(hash, 1);
551 fhdr->identification = htonl(id);
552}
553
554int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) 540int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
555{ 541{
556 struct sk_buff *frag; 542 struct sk_buff *frag;
diff --git a/net/ipv6/output_core.c b/net/ipv6/output_core.c
index 97f41a3e68d9..54520a0bd5e3 100644
--- a/net/ipv6/output_core.c
+++ b/net/ipv6/output_core.c
@@ -9,6 +9,24 @@
9#include <net/addrconf.h> 9#include <net/addrconf.h>
10#include <net/secure_seq.h> 10#include <net/secure_seq.h>
11 11
12u32 __ipv6_select_ident(u32 hashrnd, struct in6_addr *dst, struct in6_addr *src)
13{
14 u32 hash, id;
15
16 hash = __ipv6_addr_jhash(dst, hashrnd);
17 hash = __ipv6_addr_jhash(src, hash);
18
19 /* Treat id of 0 as unset and if we get 0 back from ip_idents_reserve,
20 * set the hight order instead thus minimizing possible future
21 * collisions.
22 */
23 id = ip_idents_reserve(hash, 1);
24 if (unlikely(!id))
25 id = 1 << 31;
26
27 return id;
28}
29
12/* This function exists only for tap drivers that must support broken 30/* This function exists only for tap drivers that must support broken
13 * clients requesting UFO without specifying an IPv6 fragment ID. 31 * clients requesting UFO without specifying an IPv6 fragment ID.
14 * 32 *
@@ -22,7 +40,7 @@ void ipv6_proxy_select_ident(struct sk_buff *skb)
22 static u32 ip6_proxy_idents_hashrnd __read_mostly; 40 static u32 ip6_proxy_idents_hashrnd __read_mostly;
23 struct in6_addr buf[2]; 41 struct in6_addr buf[2];
24 struct in6_addr *addrs; 42 struct in6_addr *addrs;
25 u32 hash, id; 43 u32 id;
26 44
27 addrs = skb_header_pointer(skb, 45 addrs = skb_header_pointer(skb,
28 skb_network_offset(skb) + 46 skb_network_offset(skb) +
@@ -34,14 +52,25 @@ void ipv6_proxy_select_ident(struct sk_buff *skb)
34 net_get_random_once(&ip6_proxy_idents_hashrnd, 52 net_get_random_once(&ip6_proxy_idents_hashrnd,
35 sizeof(ip6_proxy_idents_hashrnd)); 53 sizeof(ip6_proxy_idents_hashrnd));
36 54
37 hash = __ipv6_addr_jhash(&addrs[1], ip6_proxy_idents_hashrnd); 55 id = __ipv6_select_ident(ip6_proxy_idents_hashrnd,
38 hash = __ipv6_addr_jhash(&addrs[0], hash); 56 &addrs[1], &addrs[0]);
39 57 skb_shinfo(skb)->ip6_frag_id = id;
40 id = ip_idents_reserve(hash, 1);
41 skb_shinfo(skb)->ip6_frag_id = htonl(id);
42} 58}
43EXPORT_SYMBOL_GPL(ipv6_proxy_select_ident); 59EXPORT_SYMBOL_GPL(ipv6_proxy_select_ident);
44 60
61void ipv6_select_ident(struct frag_hdr *fhdr, struct rt6_info *rt)
62{
63 static u32 ip6_idents_hashrnd __read_mostly;
64 u32 id;
65
66 net_get_random_once(&ip6_idents_hashrnd, sizeof(ip6_idents_hashrnd));
67
68 id = __ipv6_select_ident(ip6_idents_hashrnd, &rt->rt6i_dst.addr,
69 &rt->rt6i_src.addr);
70 fhdr->identification = htonl(id);
71}
72EXPORT_SYMBOL(ipv6_select_ident);
73
45int ip6_find_1stfragopt(struct sk_buff *skb, u8 **nexthdr) 74int ip6_find_1stfragopt(struct sk_buff *skb, u8 **nexthdr)
46{ 75{
47 u16 offset = sizeof(struct ipv6hdr); 76 u16 offset = sizeof(struct ipv6hdr);
diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c
index 213546bd6d5d..cdbfe5af6187 100644
--- a/net/ipv6/sit.c
+++ b/net/ipv6/sit.c
@@ -1506,12 +1506,12 @@ static bool ipip6_netlink_encap_parms(struct nlattr *data[],
1506 1506
1507 if (data[IFLA_IPTUN_ENCAP_SPORT]) { 1507 if (data[IFLA_IPTUN_ENCAP_SPORT]) {
1508 ret = true; 1508 ret = true;
1509 ipencap->sport = nla_get_u16(data[IFLA_IPTUN_ENCAP_SPORT]); 1509 ipencap->sport = nla_get_be16(data[IFLA_IPTUN_ENCAP_SPORT]);
1510 } 1510 }
1511 1511
1512 if (data[IFLA_IPTUN_ENCAP_DPORT]) { 1512 if (data[IFLA_IPTUN_ENCAP_DPORT]) {
1513 ret = true; 1513 ret = true;
1514 ipencap->dport = nla_get_u16(data[IFLA_IPTUN_ENCAP_DPORT]); 1514 ipencap->dport = nla_get_be16(data[IFLA_IPTUN_ENCAP_DPORT]);
1515 } 1515 }
1516 1516
1517 return ret; 1517 return ret;
@@ -1707,9 +1707,9 @@ static int ipip6_fill_info(struct sk_buff *skb, const struct net_device *dev)
1707 1707
1708 if (nla_put_u16(skb, IFLA_IPTUN_ENCAP_TYPE, 1708 if (nla_put_u16(skb, IFLA_IPTUN_ENCAP_TYPE,
1709 tunnel->encap.type) || 1709 tunnel->encap.type) ||
1710 nla_put_u16(skb, IFLA_IPTUN_ENCAP_SPORT, 1710 nla_put_be16(skb, IFLA_IPTUN_ENCAP_SPORT,
1711 tunnel->encap.sport) || 1711 tunnel->encap.sport) ||
1712 nla_put_u16(skb, IFLA_IPTUN_ENCAP_DPORT, 1712 nla_put_be16(skb, IFLA_IPTUN_ENCAP_DPORT,
1713 tunnel->encap.dport) || 1713 tunnel->encap.dport) ||
1714 nla_put_u16(skb, IFLA_IPTUN_ENCAP_FLAGS, 1714 nla_put_u16(skb, IFLA_IPTUN_ENCAP_FLAGS,
1715 tunnel->encap.flags)) 1715 tunnel->encap.flags))
diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c
index b6aa8ed18257..a56276996b72 100644
--- a/net/ipv6/udp_offload.c
+++ b/net/ipv6/udp_offload.c
@@ -52,6 +52,10 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
52 52
53 skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss); 53 skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss);
54 54
55 /* Set the IPv6 fragment id if not set yet */
56 if (!skb_shinfo(skb)->ip6_frag_id)
57 ipv6_proxy_select_ident(skb);
58
55 segs = NULL; 59 segs = NULL;
56 goto out; 60 goto out;
57 } 61 }
@@ -108,7 +112,11 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
108 fptr = (struct frag_hdr *)(skb_network_header(skb) + unfrag_ip6hlen); 112 fptr = (struct frag_hdr *)(skb_network_header(skb) + unfrag_ip6hlen);
109 fptr->nexthdr = nexthdr; 113 fptr->nexthdr = nexthdr;
110 fptr->reserved = 0; 114 fptr->reserved = 0;
111 fptr->identification = skb_shinfo(skb)->ip6_frag_id; 115 if (skb_shinfo(skb)->ip6_frag_id)
116 fptr->identification = skb_shinfo(skb)->ip6_frag_id;
117 else
118 ipv6_select_ident(fptr,
119 (struct rt6_info *)skb_dst(skb));
112 120
113 /* Fragment the skb. ipv6 header and the remaining fields of the 121 /* Fragment the skb. ipv6 header and the remaining fields of the
114 * fragment header are updated in ipv6_gso_segment() 122 * fragment header are updated in ipv6_gso_segment()
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 990decba1fe4..b87ca32efa0b 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -659,16 +659,24 @@ static inline int ip_vs_gather_frags(struct sk_buff *skb, u_int32_t user)
659 return err; 659 return err;
660} 660}
661 661
662static int ip_vs_route_me_harder(int af, struct sk_buff *skb) 662static int ip_vs_route_me_harder(int af, struct sk_buff *skb,
663 unsigned int hooknum)
663{ 664{
665 if (!sysctl_snat_reroute(skb))
666 return 0;
667 /* Reroute replies only to remote clients (FORWARD and LOCAL_OUT) */
668 if (NF_INET_LOCAL_IN == hooknum)
669 return 0;
664#ifdef CONFIG_IP_VS_IPV6 670#ifdef CONFIG_IP_VS_IPV6
665 if (af == AF_INET6) { 671 if (af == AF_INET6) {
666 if (sysctl_snat_reroute(skb) && ip6_route_me_harder(skb) != 0) 672 struct dst_entry *dst = skb_dst(skb);
673
674 if (dst->dev && !(dst->dev->flags & IFF_LOOPBACK) &&
675 ip6_route_me_harder(skb) != 0)
667 return 1; 676 return 1;
668 } else 677 } else
669#endif 678#endif
670 if ((sysctl_snat_reroute(skb) || 679 if (!(skb_rtable(skb)->rt_flags & RTCF_LOCAL) &&
671 skb_rtable(skb)->rt_flags & RTCF_LOCAL) &&
672 ip_route_me_harder(skb, RTN_LOCAL) != 0) 680 ip_route_me_harder(skb, RTN_LOCAL) != 0)
673 return 1; 681 return 1;
674 682
@@ -791,7 +799,8 @@ static int handle_response_icmp(int af, struct sk_buff *skb,
791 union nf_inet_addr *snet, 799 union nf_inet_addr *snet,
792 __u8 protocol, struct ip_vs_conn *cp, 800 __u8 protocol, struct ip_vs_conn *cp,
793 struct ip_vs_protocol *pp, 801 struct ip_vs_protocol *pp,
794 unsigned int offset, unsigned int ihl) 802 unsigned int offset, unsigned int ihl,
803 unsigned int hooknum)
795{ 804{
796 unsigned int verdict = NF_DROP; 805 unsigned int verdict = NF_DROP;
797 806
@@ -821,7 +830,7 @@ static int handle_response_icmp(int af, struct sk_buff *skb,
821#endif 830#endif
822 ip_vs_nat_icmp(skb, pp, cp, 1); 831 ip_vs_nat_icmp(skb, pp, cp, 1);
823 832
824 if (ip_vs_route_me_harder(af, skb)) 833 if (ip_vs_route_me_harder(af, skb, hooknum))
825 goto out; 834 goto out;
826 835
827 /* do the statistics and put it back */ 836 /* do the statistics and put it back */
@@ -916,7 +925,7 @@ static int ip_vs_out_icmp(struct sk_buff *skb, int *related,
916 925
917 snet.ip = iph->saddr; 926 snet.ip = iph->saddr;
918 return handle_response_icmp(AF_INET, skb, &snet, cih->protocol, cp, 927 return handle_response_icmp(AF_INET, skb, &snet, cih->protocol, cp,
919 pp, ciph.len, ihl); 928 pp, ciph.len, ihl, hooknum);
920} 929}
921 930
922#ifdef CONFIG_IP_VS_IPV6 931#ifdef CONFIG_IP_VS_IPV6
@@ -981,7 +990,8 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
981 snet.in6 = ciph.saddr.in6; 990 snet.in6 = ciph.saddr.in6;
982 writable = ciph.len; 991 writable = ciph.len;
983 return handle_response_icmp(AF_INET6, skb, &snet, ciph.protocol, cp, 992 return handle_response_icmp(AF_INET6, skb, &snet, ciph.protocol, cp,
984 pp, writable, sizeof(struct ipv6hdr)); 993 pp, writable, sizeof(struct ipv6hdr),
994 hooknum);
985} 995}
986#endif 996#endif
987 997
@@ -1040,7 +1050,8 @@ static inline bool is_new_conn(const struct sk_buff *skb,
1040 */ 1050 */
1041static unsigned int 1051static unsigned int
1042handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, 1052handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
1043 struct ip_vs_conn *cp, struct ip_vs_iphdr *iph) 1053 struct ip_vs_conn *cp, struct ip_vs_iphdr *iph,
1054 unsigned int hooknum)
1044{ 1055{
1045 struct ip_vs_protocol *pp = pd->pp; 1056 struct ip_vs_protocol *pp = pd->pp;
1046 1057
@@ -1078,7 +1089,7 @@ handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
1078 * if it came from this machine itself. So re-compute 1089 * if it came from this machine itself. So re-compute
1079 * the routing information. 1090 * the routing information.
1080 */ 1091 */
1081 if (ip_vs_route_me_harder(af, skb)) 1092 if (ip_vs_route_me_harder(af, skb, hooknum))
1082 goto drop; 1093 goto drop;
1083 1094
1084 IP_VS_DBG_PKT(10, af, pp, skb, 0, "After SNAT"); 1095 IP_VS_DBG_PKT(10, af, pp, skb, 0, "After SNAT");
@@ -1181,7 +1192,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
1181 cp = pp->conn_out_get(af, skb, &iph, 0); 1192 cp = pp->conn_out_get(af, skb, &iph, 0);
1182 1193
1183 if (likely(cp)) 1194 if (likely(cp))
1184 return handle_response(af, skb, pd, cp, &iph); 1195 return handle_response(af, skb, pd, cp, &iph, hooknum);
1185 if (sysctl_nat_icmp_send(net) && 1196 if (sysctl_nat_icmp_send(net) &&
1186 (pp->protocol == IPPROTO_TCP || 1197 (pp->protocol == IPPROTO_TCP ||
1187 pp->protocol == IPPROTO_UDP || 1198 pp->protocol == IPPROTO_UDP ||
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 3b3ddb4fb9ee..1ff04bcd4871 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -1134,9 +1134,11 @@ static struct nft_stats __percpu *nft_stats_alloc(const struct nlattr *attr)
1134 /* Restore old counters on this cpu, no problem. Per-cpu statistics 1134 /* Restore old counters on this cpu, no problem. Per-cpu statistics
1135 * are not exposed to userspace. 1135 * are not exposed to userspace.
1136 */ 1136 */
1137 preempt_disable();
1137 stats = this_cpu_ptr(newstats); 1138 stats = this_cpu_ptr(newstats);
1138 stats->bytes = be64_to_cpu(nla_get_be64(tb[NFTA_COUNTER_BYTES])); 1139 stats->bytes = be64_to_cpu(nla_get_be64(tb[NFTA_COUNTER_BYTES]));
1139 stats->pkts = be64_to_cpu(nla_get_be64(tb[NFTA_COUNTER_PACKETS])); 1140 stats->pkts = be64_to_cpu(nla_get_be64(tb[NFTA_COUNTER_PACKETS]));
1141 preempt_enable();
1140 1142
1141 return newstats; 1143 return newstats;
1142} 1144}
@@ -1262,8 +1264,10 @@ static int nf_tables_newchain(struct sock *nlsk, struct sk_buff *skb,
1262 nft_ctx_init(&ctx, skb, nlh, afi, table, chain, nla); 1264 nft_ctx_init(&ctx, skb, nlh, afi, table, chain, nla);
1263 trans = nft_trans_alloc(&ctx, NFT_MSG_NEWCHAIN, 1265 trans = nft_trans_alloc(&ctx, NFT_MSG_NEWCHAIN,
1264 sizeof(struct nft_trans_chain)); 1266 sizeof(struct nft_trans_chain));
1265 if (trans == NULL) 1267 if (trans == NULL) {
1268 free_percpu(stats);
1266 return -ENOMEM; 1269 return -ENOMEM;
1270 }
1267 1271
1268 nft_trans_chain_stats(trans) = stats; 1272 nft_trans_chain_stats(trans) = stats;
1269 nft_trans_chain_update(trans) = true; 1273 nft_trans_chain_update(trans) = true;
@@ -1319,8 +1323,10 @@ static int nf_tables_newchain(struct sock *nlsk, struct sk_buff *skb,
1319 hookfn = type->hooks[hooknum]; 1323 hookfn = type->hooks[hooknum];
1320 1324
1321 basechain = kzalloc(sizeof(*basechain), GFP_KERNEL); 1325 basechain = kzalloc(sizeof(*basechain), GFP_KERNEL);
1322 if (basechain == NULL) 1326 if (basechain == NULL) {
1327 module_put(type->owner);
1323 return -ENOMEM; 1328 return -ENOMEM;
1329 }
1324 1330
1325 if (nla[NFTA_CHAIN_COUNTERS]) { 1331 if (nla[NFTA_CHAIN_COUNTERS]) {
1326 stats = nft_stats_alloc(nla[NFTA_CHAIN_COUNTERS]); 1332 stats = nft_stats_alloc(nla[NFTA_CHAIN_COUNTERS]);
@@ -3753,6 +3759,24 @@ int nft_chain_validate_dependency(const struct nft_chain *chain,
3753} 3759}
3754EXPORT_SYMBOL_GPL(nft_chain_validate_dependency); 3760EXPORT_SYMBOL_GPL(nft_chain_validate_dependency);
3755 3761
3762int nft_chain_validate_hooks(const struct nft_chain *chain,
3763 unsigned int hook_flags)
3764{
3765 struct nft_base_chain *basechain;
3766
3767 if (chain->flags & NFT_BASE_CHAIN) {
3768 basechain = nft_base_chain(chain);
3769
3770 if ((1 << basechain->ops[0].hooknum) & hook_flags)
3771 return 0;
3772
3773 return -EOPNOTSUPP;
3774 }
3775
3776 return 0;
3777}
3778EXPORT_SYMBOL_GPL(nft_chain_validate_hooks);
3779
3756/* 3780/*
3757 * Loop detection - walk through the ruleset beginning at the destination chain 3781 * Loop detection - walk through the ruleset beginning at the destination chain
3758 * of a new jump until either the source chain is reached (loop) or all 3782 * of a new jump until either the source chain is reached (loop) or all
diff --git a/net/netfilter/nft_masq.c b/net/netfilter/nft_masq.c
index d1ffd5eb3a9b..9aea747b43ea 100644
--- a/net/netfilter/nft_masq.c
+++ b/net/netfilter/nft_masq.c
@@ -21,6 +21,21 @@ const struct nla_policy nft_masq_policy[NFTA_MASQ_MAX + 1] = {
21}; 21};
22EXPORT_SYMBOL_GPL(nft_masq_policy); 22EXPORT_SYMBOL_GPL(nft_masq_policy);
23 23
24int nft_masq_validate(const struct nft_ctx *ctx,
25 const struct nft_expr *expr,
26 const struct nft_data **data)
27{
28 int err;
29
30 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
31 if (err < 0)
32 return err;
33
34 return nft_chain_validate_hooks(ctx->chain,
35 (1 << NF_INET_POST_ROUTING));
36}
37EXPORT_SYMBOL_GPL(nft_masq_validate);
38
24int nft_masq_init(const struct nft_ctx *ctx, 39int nft_masq_init(const struct nft_ctx *ctx,
25 const struct nft_expr *expr, 40 const struct nft_expr *expr,
26 const struct nlattr * const tb[]) 41 const struct nlattr * const tb[])
@@ -28,8 +43,8 @@ int nft_masq_init(const struct nft_ctx *ctx,
28 struct nft_masq *priv = nft_expr_priv(expr); 43 struct nft_masq *priv = nft_expr_priv(expr);
29 int err; 44 int err;
30 45
31 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT); 46 err = nft_masq_validate(ctx, expr, NULL);
32 if (err < 0) 47 if (err)
33 return err; 48 return err;
34 49
35 if (tb[NFTA_MASQ_FLAGS] == NULL) 50 if (tb[NFTA_MASQ_FLAGS] == NULL)
@@ -60,12 +75,5 @@ nla_put_failure:
60} 75}
61EXPORT_SYMBOL_GPL(nft_masq_dump); 76EXPORT_SYMBOL_GPL(nft_masq_dump);
62 77
63int nft_masq_validate(const struct nft_ctx *ctx, const struct nft_expr *expr,
64 const struct nft_data **data)
65{
66 return nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
67}
68EXPORT_SYMBOL_GPL(nft_masq_validate);
69
70MODULE_LICENSE("GPL"); 78MODULE_LICENSE("GPL");
71MODULE_AUTHOR("Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>"); 79MODULE_AUTHOR("Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>");
diff --git a/net/netfilter/nft_nat.c b/net/netfilter/nft_nat.c
index aff54fb1c8a0..a0837c6c9283 100644
--- a/net/netfilter/nft_nat.c
+++ b/net/netfilter/nft_nat.c
@@ -88,17 +88,40 @@ static const struct nla_policy nft_nat_policy[NFTA_NAT_MAX + 1] = {
88 [NFTA_NAT_FLAGS] = { .type = NLA_U32 }, 88 [NFTA_NAT_FLAGS] = { .type = NLA_U32 },
89}; 89};
90 90
91static int nft_nat_init(const struct nft_ctx *ctx, const struct nft_expr *expr, 91static int nft_nat_validate(const struct nft_ctx *ctx,
92 const struct nlattr * const tb[]) 92 const struct nft_expr *expr,
93 const struct nft_data **data)
93{ 94{
94 struct nft_nat *priv = nft_expr_priv(expr); 95 struct nft_nat *priv = nft_expr_priv(expr);
95 u32 family;
96 int err; 96 int err;
97 97
98 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT); 98 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
99 if (err < 0) 99 if (err < 0)
100 return err; 100 return err;
101 101
102 switch (priv->type) {
103 case NFT_NAT_SNAT:
104 err = nft_chain_validate_hooks(ctx->chain,
105 (1 << NF_INET_POST_ROUTING) |
106 (1 << NF_INET_LOCAL_IN));
107 break;
108 case NFT_NAT_DNAT:
109 err = nft_chain_validate_hooks(ctx->chain,
110 (1 << NF_INET_PRE_ROUTING) |
111 (1 << NF_INET_LOCAL_OUT));
112 break;
113 }
114
115 return err;
116}
117
118static int nft_nat_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
119 const struct nlattr * const tb[])
120{
121 struct nft_nat *priv = nft_expr_priv(expr);
122 u32 family;
123 int err;
124
102 if (tb[NFTA_NAT_TYPE] == NULL || 125 if (tb[NFTA_NAT_TYPE] == NULL ||
103 (tb[NFTA_NAT_REG_ADDR_MIN] == NULL && 126 (tb[NFTA_NAT_REG_ADDR_MIN] == NULL &&
104 tb[NFTA_NAT_REG_PROTO_MIN] == NULL)) 127 tb[NFTA_NAT_REG_PROTO_MIN] == NULL))
@@ -115,6 +138,10 @@ static int nft_nat_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
115 return -EINVAL; 138 return -EINVAL;
116 } 139 }
117 140
141 err = nft_nat_validate(ctx, expr, NULL);
142 if (err < 0)
143 return err;
144
118 if (tb[NFTA_NAT_FAMILY] == NULL) 145 if (tb[NFTA_NAT_FAMILY] == NULL)
119 return -EINVAL; 146 return -EINVAL;
120 147
@@ -219,13 +246,6 @@ nla_put_failure:
219 return -1; 246 return -1;
220} 247}
221 248
222static int nft_nat_validate(const struct nft_ctx *ctx,
223 const struct nft_expr *expr,
224 const struct nft_data **data)
225{
226 return nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
227}
228
229static struct nft_expr_type nft_nat_type; 249static struct nft_expr_type nft_nat_type;
230static const struct nft_expr_ops nft_nat_ops = { 250static const struct nft_expr_ops nft_nat_ops = {
231 .type = &nft_nat_type, 251 .type = &nft_nat_type,
diff --git a/net/netfilter/nft_redir.c b/net/netfilter/nft_redir.c
index 9e8093f28311..d7e9e93a4e90 100644
--- a/net/netfilter/nft_redir.c
+++ b/net/netfilter/nft_redir.c
@@ -23,6 +23,22 @@ const struct nla_policy nft_redir_policy[NFTA_REDIR_MAX + 1] = {
23}; 23};
24EXPORT_SYMBOL_GPL(nft_redir_policy); 24EXPORT_SYMBOL_GPL(nft_redir_policy);
25 25
26int nft_redir_validate(const struct nft_ctx *ctx,
27 const struct nft_expr *expr,
28 const struct nft_data **data)
29{
30 int err;
31
32 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
33 if (err < 0)
34 return err;
35
36 return nft_chain_validate_hooks(ctx->chain,
37 (1 << NF_INET_PRE_ROUTING) |
38 (1 << NF_INET_LOCAL_OUT));
39}
40EXPORT_SYMBOL_GPL(nft_redir_validate);
41
26int nft_redir_init(const struct nft_ctx *ctx, 42int nft_redir_init(const struct nft_ctx *ctx,
27 const struct nft_expr *expr, 43 const struct nft_expr *expr,
28 const struct nlattr * const tb[]) 44 const struct nlattr * const tb[])
@@ -30,7 +46,7 @@ int nft_redir_init(const struct nft_ctx *ctx,
30 struct nft_redir *priv = nft_expr_priv(expr); 46 struct nft_redir *priv = nft_expr_priv(expr);
31 int err; 47 int err;
32 48
33 err = nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT); 49 err = nft_redir_validate(ctx, expr, NULL);
34 if (err < 0) 50 if (err < 0)
35 return err; 51 return err;
36 52
@@ -88,12 +104,5 @@ nla_put_failure:
88} 104}
89EXPORT_SYMBOL_GPL(nft_redir_dump); 105EXPORT_SYMBOL_GPL(nft_redir_dump);
90 106
91int nft_redir_validate(const struct nft_ctx *ctx, const struct nft_expr *expr,
92 const struct nft_data **data)
93{
94 return nft_chain_validate_dependency(ctx->chain, NFT_CHAIN_T_NAT);
95}
96EXPORT_SYMBOL_GPL(nft_redir_validate);
97
98MODULE_LICENSE("GPL"); 107MODULE_LICENSE("GPL");
99MODULE_AUTHOR("Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>"); 108MODULE_AUTHOR("Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>");
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 02fdde28dada..75532efa51cd 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -1438,7 +1438,7 @@ static void netlink_undo_bind(int group, long unsigned int groups,
1438 1438
1439 for (undo = 0; undo < group; undo++) 1439 for (undo = 0; undo < group; undo++)
1440 if (test_bit(undo, &groups)) 1440 if (test_bit(undo, &groups))
1441 nlk->netlink_unbind(sock_net(sk), undo); 1441 nlk->netlink_unbind(sock_net(sk), undo + 1);
1442} 1442}
1443 1443
1444static int netlink_bind(struct socket *sock, struct sockaddr *addr, 1444static int netlink_bind(struct socket *sock, struct sockaddr *addr,
@@ -1476,7 +1476,7 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr,
1476 for (group = 0; group < nlk->ngroups; group++) { 1476 for (group = 0; group < nlk->ngroups; group++) {
1477 if (!test_bit(group, &groups)) 1477 if (!test_bit(group, &groups))
1478 continue; 1478 continue;
1479 err = nlk->netlink_bind(net, group); 1479 err = nlk->netlink_bind(net, group + 1);
1480 if (!err) 1480 if (!err)
1481 continue; 1481 continue;
1482 netlink_undo_bind(group, groups, sk); 1482 netlink_undo_bind(group, groups, sk);
diff --git a/net/rds/sysctl.c b/net/rds/sysctl.c
index c3b0cd43eb56..c173f69e1479 100644
--- a/net/rds/sysctl.c
+++ b/net/rds/sysctl.c
@@ -71,14 +71,14 @@ static struct ctl_table rds_sysctl_rds_table[] = {
71 { 71 {
72 .procname = "max_unacked_packets", 72 .procname = "max_unacked_packets",
73 .data = &rds_sysctl_max_unacked_packets, 73 .data = &rds_sysctl_max_unacked_packets,
74 .maxlen = sizeof(unsigned long), 74 .maxlen = sizeof(int),
75 .mode = 0644, 75 .mode = 0644,
76 .proc_handler = proc_dointvec, 76 .proc_handler = proc_dointvec,
77 }, 77 },
78 { 78 {
79 .procname = "max_unacked_bytes", 79 .procname = "max_unacked_bytes",
80 .data = &rds_sysctl_max_unacked_bytes, 80 .data = &rds_sysctl_max_unacked_bytes,
81 .maxlen = sizeof(unsigned long), 81 .maxlen = sizeof(int),
82 .mode = 0644, 82 .mode = 0644,
83 .proc_handler = proc_dointvec, 83 .proc_handler = proc_dointvec,
84 }, 84 },
diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index aad6a679fb13..baef987fe2c0 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -556,8 +556,9 @@ void tcf_exts_change(struct tcf_proto *tp, struct tcf_exts *dst,
556} 556}
557EXPORT_SYMBOL(tcf_exts_change); 557EXPORT_SYMBOL(tcf_exts_change);
558 558
559#define tcf_exts_first_act(ext) \ 559#define tcf_exts_first_act(ext) \
560 list_first_entry(&(exts)->actions, struct tc_action, list) 560 list_first_entry_or_null(&(exts)->actions, \
561 struct tc_action, list)
561 562
562int tcf_exts_dump(struct sk_buff *skb, struct tcf_exts *exts) 563int tcf_exts_dump(struct sk_buff *skb, struct tcf_exts *exts)
563{ 564{
@@ -603,7 +604,7 @@ int tcf_exts_dump_stats(struct sk_buff *skb, struct tcf_exts *exts)
603{ 604{
604#ifdef CONFIG_NET_CLS_ACT 605#ifdef CONFIG_NET_CLS_ACT
605 struct tc_action *a = tcf_exts_first_act(exts); 606 struct tc_action *a = tcf_exts_first_act(exts);
606 if (tcf_action_copy_stats(skb, a, 1) < 0) 607 if (a != NULL && tcf_action_copy_stats(skb, a, 1) < 0)
607 return -1; 608 return -1;
608#endif 609#endif
609 return 0; 610 return 0;
diff --git a/net/sched/sch_fq.c b/net/sched/sch_fq.c
index 9b05924cc386..333cd94ba381 100644
--- a/net/sched/sch_fq.c
+++ b/net/sched/sch_fq.c
@@ -670,8 +670,14 @@ static int fq_change(struct Qdisc *sch, struct nlattr *opt)
670 if (tb[TCA_FQ_FLOW_PLIMIT]) 670 if (tb[TCA_FQ_FLOW_PLIMIT])
671 q->flow_plimit = nla_get_u32(tb[TCA_FQ_FLOW_PLIMIT]); 671 q->flow_plimit = nla_get_u32(tb[TCA_FQ_FLOW_PLIMIT]);
672 672
673 if (tb[TCA_FQ_QUANTUM]) 673 if (tb[TCA_FQ_QUANTUM]) {
674 q->quantum = nla_get_u32(tb[TCA_FQ_QUANTUM]); 674 u32 quantum = nla_get_u32(tb[TCA_FQ_QUANTUM]);
675
676 if (quantum > 0)
677 q->quantum = quantum;
678 else
679 err = -EINVAL;
680 }
675 681
676 if (tb[TCA_FQ_INITIAL_QUANTUM]) 682 if (tb[TCA_FQ_INITIAL_QUANTUM])
677 q->initial_quantum = nla_get_u32(tb[TCA_FQ_INITIAL_QUANTUM]); 683 q->initial_quantum = nla_get_u32(tb[TCA_FQ_INITIAL_QUANTUM]);
diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c
index e49e231cef52..06320c8c1c86 100644
--- a/net/sctp/sm_make_chunk.c
+++ b/net/sctp/sm_make_chunk.c
@@ -2608,7 +2608,7 @@ do_addr_param:
2608 2608
2609 addr_param = param.v + sizeof(sctp_addip_param_t); 2609 addr_param = param.v + sizeof(sctp_addip_param_t);
2610 2610
2611 af = sctp_get_af_specific(param_type2af(param.p->type)); 2611 af = sctp_get_af_specific(param_type2af(addr_param->p.type));
2612 if (af == NULL) 2612 if (af == NULL)
2613 break; 2613 break;
2614 2614