aboutsummaryrefslogtreecommitdiffstats
path: root/net
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2008-06-13 10:34:47 -0400
committerLinus Torvalds <torvalds@linux-foundation.org>2008-06-13 10:34:47 -0400
commit51558576ead54c1047e4d41440626e4f9aa015ea (patch)
treeec1d609d4800dbda8c918d3142f6727dd4c2f462 /net
parentd36e311070ee3a378a54142a168ff5cfedba33d5 (diff)
parentec0a196626bd12e0ba108d7daa6d95a4fb25c2c5 (diff)
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: tcp: Revert 'process defer accept as established' changes. ipv6: Fix duplicate initialization of rawv6_prot.destroy bnx2x: Updating the Maintainer net: Eliminate flush_scheduled_work() calls while RTNL is held. drivers/net/r6040.c: correct bad use of round_jiffies() fec_mpc52xx: MPC52xx_MESSAGES_DEFAULT: 2nd NETIF_MSG_IFDOWN => IFUP ipg: fix receivemode IPG_RM_RECEIVEMULTICAST{,HASH} in ipg_nic_set_multicast_list() netfilter: nf_conntrack: fix ctnetlink related crash in nf_nat_setup_info() netfilter: Make nflog quiet when no one listen in userspace. ipv6: Fail with appropriate error code when setting not-applicable sockopt. ipv6: Check IPV6_MULTICAST_LOOP option value. ipv6: Check the hop limit setting in ancillary data. ipv6 route: Fix route lifetime in netlink message. ipv6 mcast: Check address family of gf_group in getsockopt(MS_FILTER). dccp: Bug in initial acknowledgment number assignment dccp ccid-3: X truncated due to type conversion dccp ccid-3: TFRC reverse-lookup Bug-Fix dccp ccid-2: Bug-Fix - Ack Vectors need to be ignored on request sockets dccp: Fix sparse warnings dccp ccid-3: Bug-Fix - Zero RTT is possible
Diffstat (limited to 'net')
-rw-r--r--net/dccp/ackvec.c29
-rw-r--r--net/dccp/ccids/ccid3.c14
-rw-r--r--net/dccp/ccids/lib/tfrc.c8
-rw-r--r--net/dccp/ccids/lib/tfrc.h25
-rw-r--r--net/dccp/ccids/lib/tfrc_equation.c8
-rw-r--r--net/dccp/minisocks.c8
-rw-r--r--net/dccp/options.c4
-rw-r--r--net/dccp/output.c2
-rw-r--r--net/dccp/probe.c2
-rw-r--r--net/ipv4/inet_connection_sock.c11
-rw-r--r--net/ipv4/tcp.c18
-rw-r--r--net/ipv4/tcp_input.c45
-rw-r--r--net/ipv4/tcp_ipv4.c8
-rw-r--r--net/ipv4/tcp_minisocks.c32
-rw-r--r--net/ipv4/tcp_timer.c5
-rw-r--r--net/ipv6/datagram.c5
-rw-r--r--net/ipv6/ipv6_sockglue.c12
-rw-r--r--net/ipv6/raw.c4
-rw-r--r--net/ipv6/route.c8
-rw-r--r--net/netfilter/nf_conntrack_core.c3
-rw-r--r--net/netfilter/nf_log.c4
21 files changed, 94 insertions, 161 deletions
diff --git a/net/dccp/ackvec.c b/net/dccp/ackvec.c
index 6de4bd195d28..1e8be246ad15 100644
--- a/net/dccp/ackvec.c
+++ b/net/dccp/ackvec.c
@@ -290,12 +290,12 @@ int dccp_ackvec_add(struct dccp_ackvec *av, const struct sock *sk,
290 290
291 while (1) { 291 while (1) {
292 const u8 len = dccp_ackvec_len(av, index); 292 const u8 len = dccp_ackvec_len(av, index);
293 const u8 state = dccp_ackvec_state(av, index); 293 const u8 av_state = dccp_ackvec_state(av, index);
294 /* 294 /*
295 * valid packets not yet in av_buf have a reserved 295 * valid packets not yet in av_buf have a reserved
296 * entry, with a len equal to 0. 296 * entry, with a len equal to 0.
297 */ 297 */
298 if (state == DCCP_ACKVEC_STATE_NOT_RECEIVED && 298 if (av_state == DCCP_ACKVEC_STATE_NOT_RECEIVED &&
299 len == 0 && delta == 0) { /* Found our 299 len == 0 && delta == 0) { /* Found our
300 reserved seat! */ 300 reserved seat! */
301 dccp_pr_debug("Found %llu reserved seat!\n", 301 dccp_pr_debug("Found %llu reserved seat!\n",
@@ -325,31 +325,6 @@ out_duplicate:
325 return -EILSEQ; 325 return -EILSEQ;
326} 326}
327 327
328#ifdef CONFIG_IP_DCCP_DEBUG
329void dccp_ackvector_print(const u64 ackno, const unsigned char *vector, int len)
330{
331 dccp_pr_debug_cat("ACK vector len=%d, ackno=%llu |", len,
332 (unsigned long long)ackno);
333
334 while (len--) {
335 const u8 state = (*vector & DCCP_ACKVEC_STATE_MASK) >> 6;
336 const u8 rl = *vector & DCCP_ACKVEC_LEN_MASK;
337
338 dccp_pr_debug_cat("%d,%d|", state, rl);
339 ++vector;
340 }
341
342 dccp_pr_debug_cat("\n");
343}
344
345void dccp_ackvec_print(const struct dccp_ackvec *av)
346{
347 dccp_ackvector_print(av->av_buf_ackno,
348 av->av_buf + av->av_buf_head,
349 av->av_vec_len);
350}
351#endif
352
353static void dccp_ackvec_throw_record(struct dccp_ackvec *av, 328static void dccp_ackvec_throw_record(struct dccp_ackvec *av,
354 struct dccp_ackvec_record *avr) 329 struct dccp_ackvec_record *avr)
355{ 330{
diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c
index f813077234b7..a1929f33d703 100644
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -159,8 +159,8 @@ static void ccid3_hc_tx_update_x(struct sock *sk, ktime_t *stamp)
159 } else if (ktime_us_delta(now, hctx->ccid3hctx_t_ld) 159 } else if (ktime_us_delta(now, hctx->ccid3hctx_t_ld)
160 - (s64)hctx->ccid3hctx_rtt >= 0) { 160 - (s64)hctx->ccid3hctx_rtt >= 0) {
161 161
162 hctx->ccid3hctx_x = 162 hctx->ccid3hctx_x = min(2 * hctx->ccid3hctx_x, min_rate);
163 max(min(2 * hctx->ccid3hctx_x, min_rate), 163 hctx->ccid3hctx_x = max(hctx->ccid3hctx_x,
164 scaled_div(((__u64)hctx->ccid3hctx_s) << 6, 164 scaled_div(((__u64)hctx->ccid3hctx_s) << 6,
165 hctx->ccid3hctx_rtt)); 165 hctx->ccid3hctx_rtt));
166 hctx->ccid3hctx_t_ld = now; 166 hctx->ccid3hctx_t_ld = now;
@@ -329,8 +329,14 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb)
329 hctx->ccid3hctx_x = rfc3390_initial_rate(sk); 329 hctx->ccid3hctx_x = rfc3390_initial_rate(sk);
330 hctx->ccid3hctx_t_ld = now; 330 hctx->ccid3hctx_t_ld = now;
331 } else { 331 } else {
332 /* Sender does not have RTT sample: X_pps = 1 pkt/sec */ 332 /*
333 hctx->ccid3hctx_x = hctx->ccid3hctx_s; 333 * Sender does not have RTT sample:
334 * - set fallback RTT (RFC 4340, 3.4) since a RTT value
335 * is needed in several parts (e.g. window counter);
336 * - set sending rate X_pps = 1pps as per RFC 3448, 4.2.
337 */
338 hctx->ccid3hctx_rtt = DCCP_FALLBACK_RTT;
339 hctx->ccid3hctx_x = hctx->ccid3hctx_s;
334 hctx->ccid3hctx_x <<= 6; 340 hctx->ccid3hctx_x <<= 6;
335 } 341 }
336 ccid3_update_send_interval(hctx); 342 ccid3_update_send_interval(hctx);
diff --git a/net/dccp/ccids/lib/tfrc.c b/net/dccp/ccids/lib/tfrc.c
index d1dfbb8de64c..97ecec0a8e76 100644
--- a/net/dccp/ccids/lib/tfrc.c
+++ b/net/dccp/ccids/lib/tfrc.c
@@ -14,14 +14,6 @@ module_param(tfrc_debug, bool, 0444);
14MODULE_PARM_DESC(tfrc_debug, "Enable debug messages"); 14MODULE_PARM_DESC(tfrc_debug, "Enable debug messages");
15#endif 15#endif
16 16
17extern int tfrc_tx_packet_history_init(void);
18extern void tfrc_tx_packet_history_exit(void);
19extern int tfrc_rx_packet_history_init(void);
20extern void tfrc_rx_packet_history_exit(void);
21
22extern int tfrc_li_init(void);
23extern void tfrc_li_exit(void);
24
25static int __init tfrc_module_init(void) 17static int __init tfrc_module_init(void)
26{ 18{
27 int rc = tfrc_li_init(); 19 int rc = tfrc_li_init();
diff --git a/net/dccp/ccids/lib/tfrc.h b/net/dccp/ccids/lib/tfrc.h
index 1fb1187bbf1c..ed9857527acf 100644
--- a/net/dccp/ccids/lib/tfrc.h
+++ b/net/dccp/ccids/lib/tfrc.h
@@ -15,7 +15,7 @@
15 * (at your option) any later version. 15 * (at your option) any later version.
16 */ 16 */
17#include <linux/types.h> 17#include <linux/types.h>
18#include <asm/div64.h> 18#include <linux/math64.h>
19#include "../../dccp.h" 19#include "../../dccp.h"
20/* internal includes that this module exports: */ 20/* internal includes that this module exports: */
21#include "loss_interval.h" 21#include "loss_interval.h"
@@ -29,21 +29,19 @@ extern int tfrc_debug;
29#endif 29#endif
30 30
31/* integer-arithmetic divisions of type (a * 1000000)/b */ 31/* integer-arithmetic divisions of type (a * 1000000)/b */
32static inline u64 scaled_div(u64 a, u32 b) 32static inline u64 scaled_div(u64 a, u64 b)
33{ 33{
34 BUG_ON(b==0); 34 BUG_ON(b==0);
35 a *= 1000000; 35 return div64_u64(a * 1000000, b);
36 do_div(a, b);
37 return a;
38} 36}
39 37
40static inline u32 scaled_div32(u64 a, u32 b) 38static inline u32 scaled_div32(u64 a, u64 b)
41{ 39{
42 u64 result = scaled_div(a, b); 40 u64 result = scaled_div(a, b);
43 41
44 if (result > UINT_MAX) { 42 if (result > UINT_MAX) {
45 DCCP_CRIT("Overflow: a(%llu)/b(%u) > ~0U", 43 DCCP_CRIT("Overflow: %llu/%llu > UINT_MAX",
46 (unsigned long long)a, b); 44 (unsigned long long)a, (unsigned long long)b);
47 return UINT_MAX; 45 return UINT_MAX;
48 } 46 }
49 return result; 47 return result;
@@ -58,7 +56,14 @@ static inline u32 tfrc_ewma(const u32 avg, const u32 newval, const u8 weight)
58 return avg ? (weight * avg + (10 - weight) * newval) / 10 : newval; 56 return avg ? (weight * avg + (10 - weight) * newval) / 10 : newval;
59} 57}
60 58
61extern u32 tfrc_calc_x(u16 s, u32 R, u32 p); 59extern u32 tfrc_calc_x(u16 s, u32 R, u32 p);
62extern u32 tfrc_calc_x_reverse_lookup(u32 fvalue); 60extern u32 tfrc_calc_x_reverse_lookup(u32 fvalue);
63 61
62extern int tfrc_tx_packet_history_init(void);
63extern void tfrc_tx_packet_history_exit(void);
64extern int tfrc_rx_packet_history_init(void);
65extern void tfrc_rx_packet_history_exit(void);
66
67extern int tfrc_li_init(void);
68extern void tfrc_li_exit(void);
64#endif /* _TFRC_H_ */ 69#endif /* _TFRC_H_ */
diff --git a/net/dccp/ccids/lib/tfrc_equation.c b/net/dccp/ccids/lib/tfrc_equation.c
index e4e64b76c10c..2f20a29cffe4 100644
--- a/net/dccp/ccids/lib/tfrc_equation.c
+++ b/net/dccp/ccids/lib/tfrc_equation.c
@@ -661,7 +661,7 @@ u32 tfrc_calc_x(u16 s, u32 R, u32 p)
661 661
662EXPORT_SYMBOL_GPL(tfrc_calc_x); 662EXPORT_SYMBOL_GPL(tfrc_calc_x);
663 663
664/* 664/**
665 * tfrc_calc_x_reverse_lookup - try to find p given f(p) 665 * tfrc_calc_x_reverse_lookup - try to find p given f(p)
666 * 666 *
667 * @fvalue: function value to match, scaled by 1000000 667 * @fvalue: function value to match, scaled by 1000000
@@ -676,11 +676,11 @@ u32 tfrc_calc_x_reverse_lookup(u32 fvalue)
676 676
677 /* Error cases. */ 677 /* Error cases. */
678 if (fvalue < tfrc_calc_x_lookup[0][1]) { 678 if (fvalue < tfrc_calc_x_lookup[0][1]) {
679 DCCP_WARN("fvalue %d smaller than resolution\n", fvalue); 679 DCCP_WARN("fvalue %u smaller than resolution\n", fvalue);
680 return tfrc_calc_x_lookup[0][1]; 680 return TFRC_SMALLEST_P;
681 } 681 }
682 if (fvalue > tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][0]) { 682 if (fvalue > tfrc_calc_x_lookup[TFRC_CALC_X_ARRSIZE - 1][0]) {
683 DCCP_WARN("fvalue %d exceeds bounds!\n", fvalue); 683 DCCP_WARN("fvalue %u exceeds bounds!\n", fvalue);
684 return 1000000; 684 return 1000000;
685 } 685 }
686 686
diff --git a/net/dccp/minisocks.c b/net/dccp/minisocks.c
index 33ad48321b08..66dca5bba858 100644
--- a/net/dccp/minisocks.c
+++ b/net/dccp/minisocks.c
@@ -165,12 +165,12 @@ out_free:
165 /* See dccp_v4_conn_request */ 165 /* See dccp_v4_conn_request */
166 newdmsk->dccpms_sequence_window = req->rcv_wnd; 166 newdmsk->dccpms_sequence_window = req->rcv_wnd;
167 167
168 newdp->dccps_gar = newdp->dccps_isr = dreq->dreq_isr; 168 newdp->dccps_gar = newdp->dccps_iss = dreq->dreq_iss;
169 dccp_update_gsr(newsk, dreq->dreq_isr);
170
171 newdp->dccps_iss = dreq->dreq_iss;
172 dccp_update_gss(newsk, dreq->dreq_iss); 169 dccp_update_gss(newsk, dreq->dreq_iss);
173 170
171 newdp->dccps_isr = dreq->dreq_isr;
172 dccp_update_gsr(newsk, dreq->dreq_isr);
173
174 /* 174 /*
175 * SWL and AWL are initially adjusted so that they are not less than 175 * SWL and AWL are initially adjusted so that they are not less than
176 * the initial Sequence Numbers received and sent, respectively: 176 * the initial Sequence Numbers received and sent, respectively:
diff --git a/net/dccp/options.c b/net/dccp/options.c
index d2a84a2fecee..43bc24e761d0 100644
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -107,9 +107,11 @@ int dccp_parse_options(struct sock *sk, struct dccp_request_sock *dreq,
107 * 107 *
108 * CCID-specific options are ignored during connection setup, as 108 * CCID-specific options are ignored during connection setup, as
109 * negotiation may still be in progress (see RFC 4340, 10.3). 109 * negotiation may still be in progress (see RFC 4340, 10.3).
110 * The same applies to Ack Vectors, as these depend on the CCID.
110 * 111 *
111 */ 112 */
112 if (dreq != NULL && opt >= 128) 113 if (dreq != NULL && (opt >= 128 ||
114 opt == DCCPO_ACK_VECTOR_0 || opt == DCCPO_ACK_VECTOR_1))
113 goto ignore_option; 115 goto ignore_option;
114 116
115 switch (opt) { 117 switch (opt) {
diff --git a/net/dccp/output.c b/net/dccp/output.c
index 1f8a9b64c083..fe20068c5d8e 100644
--- a/net/dccp/output.c
+++ b/net/dccp/output.c
@@ -508,6 +508,7 @@ void dccp_send_ack(struct sock *sk)
508 508
509EXPORT_SYMBOL_GPL(dccp_send_ack); 509EXPORT_SYMBOL_GPL(dccp_send_ack);
510 510
511#if 0
511/* FIXME: Is this still necessary (11.3) - currently nowhere used by DCCP. */ 512/* FIXME: Is this still necessary (11.3) - currently nowhere used by DCCP. */
512void dccp_send_delayed_ack(struct sock *sk) 513void dccp_send_delayed_ack(struct sock *sk)
513{ 514{
@@ -538,6 +539,7 @@ void dccp_send_delayed_ack(struct sock *sk)
538 icsk->icsk_ack.timeout = timeout; 539 icsk->icsk_ack.timeout = timeout;
539 sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout); 540 sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout);
540} 541}
542#endif
541 543
542void dccp_send_sync(struct sock *sk, const u64 ackno, 544void dccp_send_sync(struct sock *sk, const u64 ackno,
543 const enum dccp_pkt_type pkt_type) 545 const enum dccp_pkt_type pkt_type)
diff --git a/net/dccp/probe.c b/net/dccp/probe.c
index 0bcdc9250279..81368a7f5379 100644
--- a/net/dccp/probe.c
+++ b/net/dccp/probe.c
@@ -42,7 +42,7 @@ static int bufsize = 64 * 1024;
42 42
43static const char procname[] = "dccpprobe"; 43static const char procname[] = "dccpprobe";
44 44
45struct { 45static struct {
46 struct kfifo *fifo; 46 struct kfifo *fifo;
47 spinlock_t lock; 47 spinlock_t lock;
48 wait_queue_head_t wait; 48 wait_queue_head_t wait;
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 828ea211ff21..045e799d3e1d 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -419,7 +419,8 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
419 struct inet_connection_sock *icsk = inet_csk(parent); 419 struct inet_connection_sock *icsk = inet_csk(parent);
420 struct request_sock_queue *queue = &icsk->icsk_accept_queue; 420 struct request_sock_queue *queue = &icsk->icsk_accept_queue;
421 struct listen_sock *lopt = queue->listen_opt; 421 struct listen_sock *lopt = queue->listen_opt;
422 int thresh = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; 422 int max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries;
423 int thresh = max_retries;
423 unsigned long now = jiffies; 424 unsigned long now = jiffies;
424 struct request_sock **reqp, *req; 425 struct request_sock **reqp, *req;
425 int i, budget; 426 int i, budget;
@@ -455,6 +456,9 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
455 } 456 }
456 } 457 }
457 458
459 if (queue->rskq_defer_accept)
460 max_retries = queue->rskq_defer_accept;
461
458 budget = 2 * (lopt->nr_table_entries / (timeout / interval)); 462 budget = 2 * (lopt->nr_table_entries / (timeout / interval));
459 i = lopt->clock_hand; 463 i = lopt->clock_hand;
460 464
@@ -462,8 +466,9 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
462 reqp=&lopt->syn_table[i]; 466 reqp=&lopt->syn_table[i];
463 while ((req = *reqp) != NULL) { 467 while ((req = *reqp) != NULL) {
464 if (time_after_eq(now, req->expires)) { 468 if (time_after_eq(now, req->expires)) {
465 if (req->retrans < thresh && 469 if ((req->retrans < (inet_rsk(req)->acked ? max_retries : thresh)) &&
466 !req->rsk_ops->rtx_syn_ack(parent, req)) { 470 (inet_rsk(req)->acked ||
471 !req->rsk_ops->rtx_syn_ack(parent, req))) {
467 unsigned long timeo; 472 unsigned long timeo;
468 473
469 if (req->retrans++ == 0) 474 if (req->retrans++ == 0)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index ab66683b8043..fc54a48fde1e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2112,12 +2112,15 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
2112 break; 2112 break;
2113 2113
2114 case TCP_DEFER_ACCEPT: 2114 case TCP_DEFER_ACCEPT:
2115 if (val < 0) { 2115 icsk->icsk_accept_queue.rskq_defer_accept = 0;
2116 err = -EINVAL; 2116 if (val > 0) {
2117 } else { 2117 /* Translate value in seconds to number of
2118 if (val > MAX_TCP_ACCEPT_DEFERRED) 2118 * retransmits */
2119 val = MAX_TCP_ACCEPT_DEFERRED; 2119 while (icsk->icsk_accept_queue.rskq_defer_accept < 32 &&
2120 icsk->icsk_accept_queue.rskq_defer_accept = val; 2120 val > ((TCP_TIMEOUT_INIT / HZ) <<
2121 icsk->icsk_accept_queue.rskq_defer_accept))
2122 icsk->icsk_accept_queue.rskq_defer_accept++;
2123 icsk->icsk_accept_queue.rskq_defer_accept++;
2121 } 2124 }
2122 break; 2125 break;
2123 2126
@@ -2299,7 +2302,8 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
2299 val = (val ? : sysctl_tcp_fin_timeout) / HZ; 2302 val = (val ? : sysctl_tcp_fin_timeout) / HZ;
2300 break; 2303 break;
2301 case TCP_DEFER_ACCEPT: 2304 case TCP_DEFER_ACCEPT:
2302 val = icsk->icsk_accept_queue.rskq_defer_accept; 2305 val = !icsk->icsk_accept_queue.rskq_defer_accept ? 0 :
2306 ((TCP_TIMEOUT_INIT / HZ) << (icsk->icsk_accept_queue.rskq_defer_accept - 1));
2303 break; 2307 break;
2304 case TCP_WINDOW_CLAMP: 2308 case TCP_WINDOW_CLAMP:
2305 val = tp->window_clamp; 2309 val = tp->window_clamp;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index eba873e9b560..cad73b7dfef0 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4541,49 +4541,6 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, struct tcphdr *th)
4541 } 4541 }
4542} 4542}
4543 4543
4544static int tcp_defer_accept_check(struct sock *sk)
4545{
4546 struct tcp_sock *tp = tcp_sk(sk);
4547
4548 if (tp->defer_tcp_accept.request) {
4549 int queued_data = tp->rcv_nxt - tp->copied_seq;
4550 int hasfin = !skb_queue_empty(&sk->sk_receive_queue) ?
4551 tcp_hdr((struct sk_buff *)
4552 sk->sk_receive_queue.prev)->fin : 0;
4553
4554 if (queued_data && hasfin)
4555 queued_data--;
4556
4557 if (queued_data &&
4558 tp->defer_tcp_accept.listen_sk->sk_state == TCP_LISTEN) {
4559 if (sock_flag(sk, SOCK_KEEPOPEN)) {
4560 inet_csk_reset_keepalive_timer(sk,
4561 keepalive_time_when(tp));
4562 } else {
4563 inet_csk_delete_keepalive_timer(sk);
4564 }
4565
4566 inet_csk_reqsk_queue_add(
4567 tp->defer_tcp_accept.listen_sk,
4568 tp->defer_tcp_accept.request,
4569 sk);
4570
4571 tp->defer_tcp_accept.listen_sk->sk_data_ready(
4572 tp->defer_tcp_accept.listen_sk, 0);
4573
4574 sock_put(tp->defer_tcp_accept.listen_sk);
4575 sock_put(sk);
4576 tp->defer_tcp_accept.listen_sk = NULL;
4577 tp->defer_tcp_accept.request = NULL;
4578 } else if (hasfin ||
4579 tp->defer_tcp_accept.listen_sk->sk_state != TCP_LISTEN) {
4580 tcp_reset(sk);
4581 return -1;
4582 }
4583 }
4584 return 0;
4585}
4586
4587static int tcp_copy_to_iovec(struct sock *sk, struct sk_buff *skb, int hlen) 4544static int tcp_copy_to_iovec(struct sock *sk, struct sk_buff *skb, int hlen)
4588{ 4545{
4589 struct tcp_sock *tp = tcp_sk(sk); 4546 struct tcp_sock *tp = tcp_sk(sk);
@@ -4944,8 +4901,6 @@ step5:
4944 4901
4945 tcp_data_snd_check(sk); 4902 tcp_data_snd_check(sk);
4946 tcp_ack_snd_check(sk); 4903 tcp_ack_snd_check(sk);
4947
4948 tcp_defer_accept_check(sk);
4949 return 0; 4904 return 0;
4950 4905
4951csum_error: 4906csum_error:
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 4f8485c67d1a..97a230026e13 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1918,14 +1918,6 @@ int tcp_v4_destroy_sock(struct sock *sk)
1918 sk->sk_sndmsg_page = NULL; 1918 sk->sk_sndmsg_page = NULL;
1919 } 1919 }
1920 1920
1921 if (tp->defer_tcp_accept.request) {
1922 reqsk_free(tp->defer_tcp_accept.request);
1923 sock_put(tp->defer_tcp_accept.listen_sk);
1924 sock_put(sk);
1925 tp->defer_tcp_accept.listen_sk = NULL;
1926 tp->defer_tcp_accept.request = NULL;
1927 }
1928
1929 atomic_dec(&tcp_sockets_allocated); 1921 atomic_dec(&tcp_sockets_allocated);
1930 1922
1931 return 0; 1923 return 0;
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 019c8c16e5cc..8245247a6ceb 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -571,8 +571,10 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
571 does sequence test, SYN is truncated, and thus we consider 571 does sequence test, SYN is truncated, and thus we consider
572 it a bare ACK. 572 it a bare ACK.
573 573
574 Both ends (listening sockets) accept the new incoming 574 If icsk->icsk_accept_queue.rskq_defer_accept, we silently drop this
575 connection and try to talk to each other. 8-) 575 bare ACK. Otherwise, we create an established connection. Both
576 ends (listening sockets) accept the new incoming connection and try
577 to talk to each other. 8-)
576 578
577 Note: This case is both harmless, and rare. Possibility is about the 579 Note: This case is both harmless, and rare. Possibility is about the
578 same as us discovering intelligent life on another plant tomorrow. 580 same as us discovering intelligent life on another plant tomorrow.
@@ -640,6 +642,13 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
640 if (!(flg & TCP_FLAG_ACK)) 642 if (!(flg & TCP_FLAG_ACK))
641 return NULL; 643 return NULL;
642 644
645 /* If TCP_DEFER_ACCEPT is set, drop bare ACK. */
646 if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept &&
647 TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
648 inet_rsk(req)->acked = 1;
649 return NULL;
650 }
651
643 /* OK, ACK is valid, create big socket and 652 /* OK, ACK is valid, create big socket and
644 * feed this segment to it. It will repeat all 653 * feed this segment to it. It will repeat all
645 * the tests. THIS SEGMENT MUST MOVE SOCKET TO 654 * the tests. THIS SEGMENT MUST MOVE SOCKET TO
@@ -678,24 +687,7 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
678 inet_csk_reqsk_queue_unlink(sk, req, prev); 687 inet_csk_reqsk_queue_unlink(sk, req, prev);
679 inet_csk_reqsk_queue_removed(sk, req); 688 inet_csk_reqsk_queue_removed(sk, req);
680 689
681 if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept && 690 inet_csk_reqsk_queue_add(sk, req, child);
682 TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
683
684 /* the accept queue handling is done is est recv slow
685 * path so lets make sure to start there
686 */
687 tcp_sk(child)->pred_flags = 0;
688 sock_hold(sk);
689 sock_hold(child);
690 tcp_sk(child)->defer_tcp_accept.listen_sk = sk;
691 tcp_sk(child)->defer_tcp_accept.request = req;
692
693 inet_csk_reset_keepalive_timer(child,
694 inet_csk(sk)->icsk_accept_queue.rskq_defer_accept * HZ);
695 } else {
696 inet_csk_reqsk_queue_add(sk, req, child);
697 }
698
699 return child; 691 return child;
700 692
701 listen_overflow: 693 listen_overflow:
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 4de68cf5f2aa..63ed9d6830e7 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -489,11 +489,6 @@ static void tcp_keepalive_timer (unsigned long data)
489 goto death; 489 goto death;
490 } 490 }
491 491
492 if (tp->defer_tcp_accept.request && sk->sk_state == TCP_ESTABLISHED) {
493 tcp_send_active_reset(sk, GFP_ATOMIC);
494 goto death;
495 }
496
497 if (!sock_flag(sk, SOCK_KEEPOPEN) || sk->sk_state == TCP_CLOSE) 492 if (!sock_flag(sk, SOCK_KEEPOPEN) || sk->sk_state == TCP_CLOSE)
498 goto out; 493 goto out;
499 494
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index b9c2de84a8a2..0f0f94a40335 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -705,6 +705,11 @@ int datagram_send_ctl(struct net *net,
705 } 705 }
706 706
707 *hlimit = *(int *)CMSG_DATA(cmsg); 707 *hlimit = *(int *)CMSG_DATA(cmsg);
708 if (*hlimit < -1 || *hlimit > 0xff) {
709 err = -EINVAL;
710 goto exit_f;
711 }
712
708 break; 713 break;
709 714
710 case IPV6_TCLASS: 715 case IPV6_TCLASS:
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 26b83e512a09..c042ce19bd14 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -67,7 +67,7 @@ int ip6_ra_control(struct sock *sk, int sel, void (*destructor)(struct sock *))
67 67
68 /* RA packet may be delivered ONLY to IPPROTO_RAW socket */ 68 /* RA packet may be delivered ONLY to IPPROTO_RAW socket */
69 if (sk->sk_type != SOCK_RAW || inet_sk(sk)->num != IPPROTO_RAW) 69 if (sk->sk_type != SOCK_RAW || inet_sk(sk)->num != IPPROTO_RAW)
70 return -EINVAL; 70 return -ENOPROTOOPT;
71 71
72 new_ra = (sel>=0) ? kmalloc(sizeof(*new_ra), GFP_KERNEL) : NULL; 72 new_ra = (sel>=0) ? kmalloc(sizeof(*new_ra), GFP_KERNEL) : NULL;
73 73
@@ -446,7 +446,7 @@ done:
446 446
447 case IPV6_MULTICAST_HOPS: 447 case IPV6_MULTICAST_HOPS:
448 if (sk->sk_type == SOCK_STREAM) 448 if (sk->sk_type == SOCK_STREAM)
449 goto e_inval; 449 break;
450 if (optlen < sizeof(int)) 450 if (optlen < sizeof(int))
451 goto e_inval; 451 goto e_inval;
452 if (val > 255 || val < -1) 452 if (val > 255 || val < -1)
@@ -458,13 +458,15 @@ done:
458 case IPV6_MULTICAST_LOOP: 458 case IPV6_MULTICAST_LOOP:
459 if (optlen < sizeof(int)) 459 if (optlen < sizeof(int))
460 goto e_inval; 460 goto e_inval;
461 if (val != valbool)
462 goto e_inval;
461 np->mc_loop = valbool; 463 np->mc_loop = valbool;
462 retv = 0; 464 retv = 0;
463 break; 465 break;
464 466
465 case IPV6_MULTICAST_IF: 467 case IPV6_MULTICAST_IF:
466 if (sk->sk_type == SOCK_STREAM) 468 if (sk->sk_type == SOCK_STREAM)
467 goto e_inval; 469 break;
468 if (optlen < sizeof(int)) 470 if (optlen < sizeof(int))
469 goto e_inval; 471 goto e_inval;
470 472
@@ -860,7 +862,7 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
860 if (sk->sk_protocol != IPPROTO_UDP && 862 if (sk->sk_protocol != IPPROTO_UDP &&
861 sk->sk_protocol != IPPROTO_UDPLITE && 863 sk->sk_protocol != IPPROTO_UDPLITE &&
862 sk->sk_protocol != IPPROTO_TCP) 864 sk->sk_protocol != IPPROTO_TCP)
863 return -EINVAL; 865 return -ENOPROTOOPT;
864 if (sk->sk_state != TCP_ESTABLISHED) 866 if (sk->sk_state != TCP_ESTABLISHED)
865 return -ENOTCONN; 867 return -ENOTCONN;
866 val = sk->sk_family; 868 val = sk->sk_family;
@@ -874,6 +876,8 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
874 return -EINVAL; 876 return -EINVAL;
875 if (copy_from_user(&gsf, optval, GROUP_FILTER_SIZE(0))) 877 if (copy_from_user(&gsf, optval, GROUP_FILTER_SIZE(0)))
876 return -EFAULT; 878 return -EFAULT;
879 if (gsf.gf_group.ss_family != AF_INET6)
880 return -EADDRNOTAVAIL;
877 lock_sock(sk); 881 lock_sock(sk);
878 err = ip6_mc_msfget(sk, &gsf, 882 err = ip6_mc_msfget(sk, &gsf,
879 (struct group_filter __user *)optval, optlen); 883 (struct group_filter __user *)optval, optlen);
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index 8fee9a15b2d3..3aee12310d94 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -1169,7 +1169,8 @@ static int raw6_destroy(struct sock *sk)
1169 lock_sock(sk); 1169 lock_sock(sk);
1170 ip6_flush_pending_frames(sk); 1170 ip6_flush_pending_frames(sk);
1171 release_sock(sk); 1171 release_sock(sk);
1172 return 0; 1172
1173 return inet6_destroy_sock(sk);
1173} 1174}
1174 1175
1175static int rawv6_init_sk(struct sock *sk) 1176static int rawv6_init_sk(struct sock *sk)
@@ -1200,7 +1201,6 @@ struct proto rawv6_prot = {
1200 .disconnect = udp_disconnect, 1201 .disconnect = udp_disconnect,
1201 .ioctl = rawv6_ioctl, 1202 .ioctl = rawv6_ioctl,
1202 .init = rawv6_init_sk, 1203 .init = rawv6_init_sk,
1203 .destroy = inet6_destroy_sock,
1204 .setsockopt = rawv6_setsockopt, 1204 .setsockopt = rawv6_setsockopt,
1205 .getsockopt = rawv6_getsockopt, 1205 .getsockopt = rawv6_getsockopt,
1206 .sendmsg = rawv6_sendmsg, 1206 .sendmsg = rawv6_sendmsg,
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 220cffe9e63b..d1f3e19b06c7 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -2196,8 +2196,12 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
2196 2196
2197 NLA_PUT_U32(skb, RTA_PRIORITY, rt->rt6i_metric); 2197 NLA_PUT_U32(skb, RTA_PRIORITY, rt->rt6i_metric);
2198 2198
2199 expires = (rt->rt6i_flags & RTF_EXPIRES) ? 2199 if (!(rt->rt6i_flags & RTF_EXPIRES))
2200 rt->rt6i_expires - jiffies : 0; 2200 expires = 0;
2201 else if (rt->rt6i_expires - jiffies < INT_MAX)
2202 expires = rt->rt6i_expires - jiffies;
2203 else
2204 expires = INT_MAX;
2201 2205
2202 if (rtnl_put_cacheinfo(skb, &rt->u.dst, 0, 0, 0, 2206 if (rtnl_put_cacheinfo(skb, &rt->u.dst, 0, 0, 0,
2203 expires, rt->u.dst.error) < 0) 2207 expires, rt->u.dst.error) < 0)
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index c4b1799da5d7..662c1ccfee26 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -196,8 +196,6 @@ destroy_conntrack(struct nf_conntrack *nfct)
196 if (l4proto && l4proto->destroy) 196 if (l4proto && l4proto->destroy)
197 l4proto->destroy(ct); 197 l4proto->destroy(ct);
198 198
199 nf_ct_ext_destroy(ct);
200
201 rcu_read_unlock(); 199 rcu_read_unlock();
202 200
203 spin_lock_bh(&nf_conntrack_lock); 201 spin_lock_bh(&nf_conntrack_lock);
@@ -520,6 +518,7 @@ static void nf_conntrack_free_rcu(struct rcu_head *head)
520 518
521void nf_conntrack_free(struct nf_conn *ct) 519void nf_conntrack_free(struct nf_conn *ct)
522{ 520{
521 nf_ct_ext_destroy(ct);
523 call_rcu(&ct->rcu, nf_conntrack_free_rcu); 522 call_rcu(&ct->rcu, nf_conntrack_free_rcu);
524} 523}
525EXPORT_SYMBOL_GPL(nf_conntrack_free); 524EXPORT_SYMBOL_GPL(nf_conntrack_free);
diff --git a/net/netfilter/nf_log.c b/net/netfilter/nf_log.c
index bc11d7092032..9fda6ee95a31 100644
--- a/net/netfilter/nf_log.c
+++ b/net/netfilter/nf_log.c
@@ -92,10 +92,6 @@ void nf_log_packet(int pf,
92 vsnprintf(prefix, sizeof(prefix), fmt, args); 92 vsnprintf(prefix, sizeof(prefix), fmt, args);
93 va_end(args); 93 va_end(args);
94 logger->logfn(pf, hooknum, skb, in, out, loginfo, prefix); 94 logger->logfn(pf, hooknum, skb, in, out, loginfo, prefix);
95 } else if (net_ratelimit()) {
96 printk(KERN_WARNING "nf_log_packet: can\'t log since "
97 "no backend logging module loaded in! Please either "
98 "load one, or disable logging explicitly\n");
99 } 95 }
100 rcu_read_unlock(); 96 rcu_read_unlock();
101} 97}