Blame view
net/ipv4/tcp_minisocks.c
26.2 KB
457c89965
|
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c
|
2 3 4 5 6 7 8 |
/* * INET An implementation of the TCP/IP protocol suite for the LINUX * operating system. INET is implemented using the BSD Socket * interface as the means of communication with the user level. * * Implementation of the Transmission Control Protocol(TCP). * |
02c30a84e
|
9 |
* Authors: Ross Biro |
1da177e4c
|
10 11 12 13 14 15 16 17 18 19 20 |
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG> * Mark Evans, <evansmp@uhura.aston.ac.uk> * Corey Minyard <wf-rch!minyard@relay.EU.net> * Florian La Roche, <flla@stud.uni-sb.de> * Charles Hedrick, <hedrick@klinzhai.rutgers.edu> * Linus Torvalds, <torvalds@cs.helsinki.fi> * Alan Cox, <gw4pts@gw4pts.ampr.org> * Matthew Dillon, <dillon@apollo.west.oic.com> * Arnt Gulbrandsen, <agulbra@nvg.unit.no> * Jorge Cwik, <jorge@laser.satlink.net> */ |
1da177e4c
|
21 22 |
#include <linux/mm.h> #include <linux/module.h> |
5a0e3ad6a
|
23 |
#include <linux/slab.h> |
1da177e4c
|
24 25 |
#include <linux/sysctl.h> #include <linux/workqueue.h> |
60e2a7780
|
26 |
#include <linux/static_key.h> |
1da177e4c
|
27 28 29 |
#include <net/tcp.h> #include <net/inet_common.h> #include <net/xfrm.h> |
e5907459c
|
30 |
#include <net/busy_poll.h> |
1da177e4c
|
31 |
|
a2a385d62
|
32 |
static bool tcp_in_window(u32 seq, u32 end_seq, u32 s_win, u32 e_win) |
1da177e4c
|
33 34 |
{ if (seq == s_win) |
a2a385d62
|
35 |
return true; |
1da177e4c
|
36 |
if (after(end_seq, s_win) && before(seq, e_win)) |
a2a385d62
|
37 |
return true; |
a02cec215
|
38 |
return seq == e_win && seq == end_seq; |
1da177e4c
|
39 |
} |
4fb17a609
|
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 |
static enum tcp_tw_status tcp_timewait_check_oow_rate_limit(struct inet_timewait_sock *tw, const struct sk_buff *skb, int mib_idx) { struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); if (!tcp_oow_rate_limited(twsk_net(tw), skb, mib_idx, &tcptw->tw_last_oow_ack_time)) { /* Send ACK. Note, we do not put the bucket, * it will be released by caller. */ return TCP_TW_ACK; } /* We are rate-limiting, so just release the tw sock and drop skb. */ inet_twsk_put(tw); return TCP_TW_SUCCESS; } |
e905a9eda
|
58 |
/* |
1da177e4c
|
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 |
* * Main purpose of TIME-WAIT state is to close connection gracefully, * when one of ends sits in LAST-ACK or CLOSING retransmitting FIN * (and, probably, tail of data) and one or more our ACKs are lost. * * What is TIME-WAIT timeout? It is associated with maximal packet * lifetime in the internet, which results in wrong conclusion, that * it is set to catch "old duplicate segments" wandering out of their path. * It is not quite correct. This timeout is calculated so that it exceeds * maximal retransmission timeout enough to allow to lose one (or more) * segments sent by peer and our ACKs. This time may be calculated from RTO. * * When TIME-WAIT socket receives RST, it means that another end * finally closed and we are allowed to kill TIME-WAIT too. * * Second purpose of TIME-WAIT is catching old duplicate segments. * Well, certainly it is pure paranoia, but if we load TIME-WAIT * with this semantics, we MUST NOT kill TIME-WAIT state with RSTs. * * If we invented some more clever way to catch duplicates * (f.e. based on PAWS), we could truncate TIME-WAIT to several RTOs. * * The algorithm below is based on FORMAL INTERPRETATION of RFCs. * When you compare it to RFCs, please, read section SEGMENT ARRIVES * from the very beginning. * * NOTE. With recycling (and later with fin-wait-2) TW bucket * is _not_ stateless. It means, that strictly speaking we must * spinlock it. I do not want! Well, probability of misbehaviour * is ridiculously low and, seems, we could use some mb() tricks * to avoid misread sequence numbers, states etc. --ANK |
4308fc58d
|
85 86 |
* * We don't need to initialize tmp_out.sack_ok as we don't use the results |
1da177e4c
|
87 88 |
*/ enum tcp_tw_status |
8feaf0c0a
|
89 90 |
tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb, const struct tcphdr *th) |
1da177e4c
|
91 92 |
{ struct tcp_options_received tmp_opt; |
4957faade
|
93 |
struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); |
a2a385d62
|
94 |
bool paws_reject = false; |
1da177e4c
|
95 |
|
bb5b7c112
|
96 |
tmp_opt.saw_tstamp = 0; |
8feaf0c0a
|
97 |
if (th->doff > (sizeof(*th) >> 2) && tcptw->tw_ts_recent_stamp) { |
eed29f17f
|
98 |
tcp_parse_options(twsk_net(tw), skb, &tmp_opt, 0, NULL); |
1da177e4c
|
99 100 |
if (tmp_opt.saw_tstamp) { |
eee2faabc
|
101 102 |
if (tmp_opt.rcv_tsecr) tmp_opt.rcv_tsecr -= tcptw->tw_ts_offset; |
8feaf0c0a
|
103 104 |
tmp_opt.ts_recent = tcptw->tw_ts_recent; tmp_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp; |
c887e6d2d
|
105 |
paws_reject = tcp_paws_reject(&tmp_opt, th->rst); |
1da177e4c
|
106 107 108 109 110 111 112 113 114 |
} } if (tw->tw_substate == TCP_FIN_WAIT2) { /* Just repeat all the checks of tcp_rcv_state_process() */ /* Out of window, send ACK */ if (paws_reject || !tcp_in_window(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq, |
8feaf0c0a
|
115 116 |
tcptw->tw_rcv_nxt, tcptw->tw_rcv_nxt + tcptw->tw_rcv_wnd)) |
4fb17a609
|
117 118 |
return tcp_timewait_check_oow_rate_limit( tw, skb, LINUX_MIB_TCPACKSKIPPEDFINWAIT2); |
1da177e4c
|
119 120 121 |
if (th->rst) goto kill; |
8feaf0c0a
|
122 |
if (th->syn && !before(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt)) |
271c3b9b7
|
123 |
return TCP_TW_RST; |
1da177e4c
|
124 125 |
/* Dup ACK? */ |
1ac530b35
|
126 127 |
if (!th->ack || !after(TCP_SKB_CB(skb)->end_seq, tcptw->tw_rcv_nxt) || |
1da177e4c
|
128 |
TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) { |
8feaf0c0a
|
129 |
inet_twsk_put(tw); |
1da177e4c
|
130 131 132 133 134 135 136 |
return TCP_TW_SUCCESS; } /* New data or FIN. If new data arrive after half-duplex close, * reset. */ if (!th->fin || |
271c3b9b7
|
137 |
TCP_SKB_CB(skb)->end_seq != tcptw->tw_rcv_nxt + 1) |
1da177e4c
|
138 |
return TCP_TW_RST; |
1da177e4c
|
139 140 |
/* FIN arrived, enter true time-wait state. */ |
8feaf0c0a
|
141 142 |
tw->tw_substate = TCP_TIME_WAIT; tcptw->tw_rcv_nxt = TCP_SKB_CB(skb)->end_seq; |
1da177e4c
|
143 |
if (tmp_opt.saw_tstamp) { |
cca9bab1b
|
144 |
tcptw->tw_ts_recent_stamp = ktime_get_seconds(); |
8feaf0c0a
|
145 |
tcptw->tw_ts_recent = tmp_opt.rcv_tsval; |
1da177e4c
|
146 |
} |
d82bae12d
|
147 |
inet_twsk_reschedule(tw, TCP_TIMEWAIT_LEN); |
1da177e4c
|
148 149 150 151 152 153 154 155 156 157 |
return TCP_TW_ACK; } /* * Now real TIME-WAIT state. * * RFC 1122: * "When a connection is [...] on TIME-WAIT state [...] * [a TCP] MAY accept a new SYN from the remote TCP to * reopen the connection directly, if it: |
e905a9eda
|
158 |
* |
1da177e4c
|
159 160 161 162 163 |
* (1) assigns its initial sequence number for the new * connection to be larger than the largest sequence * number it used on the previous connection incarnation, * and * |
e905a9eda
|
164 |
* (2) returns to TIME-WAIT state if the SYN turns out |
1da177e4c
|
165 166 167 168 |
* to be an old duplicate". */ if (!paws_reject && |
8feaf0c0a
|
169 |
(TCP_SKB_CB(skb)->seq == tcptw->tw_rcv_nxt && |
1da177e4c
|
170 171 172 173 |
(TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq || th->rst))) { /* In window segment, it may be only reset or bare ack. */ if (th->rst) { |
caa20d9ab
|
174 |
/* This is TIME_WAIT assassination, in two flavors. |
1da177e4c
|
175 176 177 |
* Oh well... nobody has a sufficient solution to this * protocol bug yet. */ |
625357aa1
|
178 |
if (twsk_net(tw)->ipv4.sysctl_tcp_rfc1337 == 0) { |
1da177e4c
|
179 |
kill: |
dbe7faa40
|
180 |
inet_twsk_deschedule_put(tw); |
1da177e4c
|
181 182 |
return TCP_TW_SUCCESS; } |
63cc357f7
|
183 184 |
} else { inet_twsk_reschedule(tw, TCP_TIMEWAIT_LEN); |
1da177e4c
|
185 |
} |
1da177e4c
|
186 187 |
if (tmp_opt.saw_tstamp) { |
8feaf0c0a
|
188 |
tcptw->tw_ts_recent = tmp_opt.rcv_tsval; |
cca9bab1b
|
189 |
tcptw->tw_ts_recent_stamp = ktime_get_seconds(); |
1da177e4c
|
190 |
} |
8feaf0c0a
|
191 |
inet_twsk_put(tw); |
1da177e4c
|
192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 |
return TCP_TW_SUCCESS; } /* Out of window segment. All the segments are ACKed immediately. The only exception is new SYN. We accept it, if it is not old duplicate and we are not in danger to be killed by delayed old duplicates. RFC check is that it has newer sequence number works at rates <40Mbit/sec. However, if paws works, it is reliable AND even more, we even may relax silly seq space cutoff. RED-PEN: we violate main RFC requirement, if this SYN will appear old duplicate (i.e. we receive RST in reply to SYN-ACK), we must return socket to time-wait state. It is not good, but not fatal yet. */ if (th->syn && !th->rst && !th->ack && !paws_reject && |
8feaf0c0a
|
213 214 215 216 |
(after(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt) || (tmp_opt.saw_tstamp && (s32)(tcptw->tw_ts_recent - tmp_opt.rcv_tsval) < 0))) { u32 isn = tcptw->tw_snd_nxt + 65535 + 2; |
1da177e4c
|
217 218 |
if (isn == 0) isn++; |
04317dafd
|
219 |
TCP_SKB_CB(skb)->tcp_tw_isn = isn; |
1da177e4c
|
220 221 222 223 |
return TCP_TW_SYN; } if (paws_reject) |
02a1d6e7a
|
224 |
__NET_INC_STATS(twsk_net(tw), LINUX_MIB_PAWSESTABREJECTED); |
1da177e4c
|
225 |
|
2de979bd7
|
226 |
if (!th->rst) { |
1da177e4c
|
227 228 229 230 231 232 233 |
/* In this case we must reset the TIMEWAIT timer. * * If it is ACKless SYN it may be both old duplicate * and new good SYN with random sequence number <rcv_nxt. * Do not reschedule in the last case. */ if (paws_reject || th->ack) |
ed2e92394
|
234 |
inet_twsk_reschedule(tw, TCP_TIMEWAIT_LEN); |
1da177e4c
|
235 |
|
4fb17a609
|
236 237 |
return tcp_timewait_check_oow_rate_limit( tw, skb, LINUX_MIB_TCPACKSKIPPEDTIMEWAIT); |
1da177e4c
|
238 |
} |
8feaf0c0a
|
239 |
inet_twsk_put(tw); |
1da177e4c
|
240 241 |
return TCP_TW_SUCCESS; } |
4bc2f18ba
|
242 |
EXPORT_SYMBOL(tcp_timewait_state_process); |
1da177e4c
|
243 |
|
e905a9eda
|
244 |
/* |
1da177e4c
|
245 |
* Move a socket to time-wait or dead fin-wait-2 state. |
e905a9eda
|
246 |
*/ |
1da177e4c
|
247 248 |
void tcp_time_wait(struct sock *sk, int state, int timeo) { |
8292a17a3
|
249 |
const struct inet_connection_sock *icsk = inet_csk(sk); |
8feaf0c0a
|
250 |
const struct tcp_sock *tp = tcp_sk(sk); |
789f558cf
|
251 |
struct inet_timewait_sock *tw; |
1946e672c
|
252 |
struct inet_timewait_death_row *tcp_death_row = &sock_net(sk)->ipv4.tcp_death_row; |
1da177e4c
|
253 |
|
1946e672c
|
254 |
tw = inet_twsk_alloc(sk, tcp_death_row, state); |
1da177e4c
|
255 |
|
00db41243
|
256 |
if (tw) { |
8feaf0c0a
|
257 |
struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); |
463c84b97
|
258 |
const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1); |
2397849ba
|
259 |
struct inet_sock *inet = inet_sk(sk); |
8feaf0c0a
|
260 |
|
2397849ba
|
261 |
tw->tw_transparent = inet->transparent; |
004836905
|
262 |
tw->tw_mark = sk->sk_mark; |
1da177e4c
|
263 |
tw->tw_rcv_wscale = tp->rx_opt.rcv_wscale; |
8feaf0c0a
|
264 265 266 267 268 |
tcptw->tw_rcv_nxt = tp->rcv_nxt; tcptw->tw_snd_nxt = tp->snd_nxt; tcptw->tw_rcv_wnd = tcp_receive_window(tp); tcptw->tw_ts_recent = tp->rx_opt.ts_recent; tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp; |
ceaa1fef6
|
269 |
tcptw->tw_ts_offset = tp->tsoffset; |
4fb17a609
|
270 |
tcptw->tw_last_oow_ack_time = 0; |
a842fe142
|
271 |
tcptw->tw_tx_delay = tp->tcp_tx_delay; |
dfd56b8b3
|
272 |
#if IS_ENABLED(CONFIG_IPV6) |
1da177e4c
|
273 274 |
if (tw->tw_family == PF_INET6) { struct ipv6_pinfo *np = inet6_sk(sk); |
efe4208f4
|
275 276 |
tw->tw_v6_daddr = sk->sk_v6_daddr; tw->tw_v6_rcv_saddr = sk->sk_v6_rcv_saddr; |
b903d324b
|
277 |
tw->tw_tclass = np->tclass; |
21858cd02
|
278 |
tw->tw_flowlabel = be32_to_cpu(np->flow_label & IPV6_FLOWLABEL_MASK); |
c67b85558
|
279 |
tw->tw_txhash = sk->sk_txhash; |
9fe516ba3
|
280 |
tw->tw_ipv6only = sk->sk_ipv6only; |
c676270bc
|
281 |
} |
1da177e4c
|
282 |
#endif |
cfb6eeb4c
|
283 284 285 286 287 288 289 290 291 |
#ifdef CONFIG_TCP_MD5SIG /* * The timewait bucket does not have the key DB from the * sock structure. We just make a quick copy of the * md5 key being used (if indeed we are using one) * so the timewait ack generating code has the key. */ do { |
a915da9b6
|
292 |
tcptw->tw_md5_key = NULL; |
6aedbf986
|
293 294 295 296 297 298 299 300 |
if (static_branch_unlikely(&tcp_md5_needed)) { struct tcp_md5sig_key *key; key = tp->af_specific->md5_lookup(sk, sk); if (key) { tcptw->tw_md5_key = kmemdup(key, sizeof(*key), GFP_ATOMIC); BUG_ON(tcptw->tw_md5_key && !tcp_alloc_md5sig_pool()); } |
cfb6eeb4c
|
301 |
} |
2de979bd7
|
302 |
} while (0); |
cfb6eeb4c
|
303 |
#endif |
1da177e4c
|
304 305 306 |
/* Get the TIME_WAIT timeout firing. */ if (timeo < rto) timeo = rto; |
d82bae12d
|
307 308 |
if (state == TCP_TIME_WAIT) timeo = TCP_TIMEWAIT_LEN; |
1da177e4c
|
309 |
|
cfac7f836
|
310 311 312 313 314 |
/* tw_timer is pinned, so we need to make sure BH are disabled * in following section, otherwise timer handler could run before * we complete the initialization. */ local_bh_disable(); |
789f558cf
|
315 |
inet_twsk_schedule(tw, timeo); |
ec94c2696
|
316 317 318 319 |
/* Linkage updates. * Note that access to tw after this point is illegal. */ inet_twsk_hashdance(tw, sk, &tcp_hashinfo); |
cfac7f836
|
320 |
local_bh_enable(); |
1da177e4c
|
321 322 323 324 325 |
} else { /* Sorry, if we're out of memory, just CLOSE this * socket up. We've got bigger problems than * non-graceful socket closings. */ |
c10d9310e
|
326 |
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPTIMEWAITOVERFLOW); |
1da177e4c
|
327 328 329 330 331 |
} tcp_update_metrics(sk); tcp_done(sk); } |
cc35c88ae
|
332 |
EXPORT_SYMBOL(tcp_time_wait); |
1da177e4c
|
333 |
|
cfb6eeb4c
|
334 335 |
void tcp_twsk_destructor(struct sock *sk) { |
b6242b9b4
|
336 |
#ifdef CONFIG_TCP_MD5SIG |
6aedbf986
|
337 338 |
if (static_branch_unlikely(&tcp_md5_needed)) { struct tcp_timewait_sock *twsk = tcp_twsk(sk); |
2397849ba
|
339 |
|
6aedbf986
|
340 341 342 |
if (twsk->tw_md5_key) kfree_rcu(twsk->tw_md5_key, rcu); } |
cfb6eeb4c
|
343 344 |
#endif } |
cfb6eeb4c
|
345 |
EXPORT_SYMBOL_GPL(tcp_twsk_destructor); |
b1964b5fc
|
346 347 348 |
/* Warning : This function is called without sk_listener being locked. * Be sure to read socket fields once, as their value could change under us. */ |
843f4a55e
|
349 |
void tcp_openreq_init_rwin(struct request_sock *req, |
b1964b5fc
|
350 351 |
const struct sock *sk_listener, const struct dst_entry *dst) |
843f4a55e
|
352 353 |
{ struct inet_request_sock *ireq = inet_rsk(req); |
b1964b5fc
|
354 |
const struct tcp_sock *tp = tcp_sk(sk_listener); |
b1964b5fc
|
355 |
int full_space = tcp_full_space(sk_listener); |
b1964b5fc
|
356 357 |
u32 window_clamp; __u8 rcv_wscale; |
13d3b1ebe
|
358 |
u32 rcv_wnd; |
3541f9e8b
|
359 |
int mss; |
843f4a55e
|
360 |
|
3541f9e8b
|
361 |
mss = tcp_mss_clamp(tp, dst_metric_advmss(dst)); |
b1964b5fc
|
362 |
window_clamp = READ_ONCE(tp->window_clamp); |
843f4a55e
|
363 |
/* Set this up on the first call only */ |
ed53d0ab7
|
364 |
req->rsk_window_clamp = window_clamp ? : dst_metric(dst, RTAX_WINDOW); |
843f4a55e
|
365 366 |
/* limit the window selection if the user enforce a smaller rx buffer */ |
b1964b5fc
|
367 |
if (sk_listener->sk_userlocks & SOCK_RCVBUF_LOCK && |
ed53d0ab7
|
368 369 |
(req->rsk_window_clamp > full_space || req->rsk_window_clamp == 0)) req->rsk_window_clamp = full_space; |
843f4a55e
|
370 |
|
13d3b1ebe
|
371 372 373 374 375 |
rcv_wnd = tcp_rwnd_init_bpf((struct sock *)req); if (rcv_wnd == 0) rcv_wnd = dst_metric(dst, RTAX_INITRWND); else if (full_space < rcv_wnd * mss) full_space = rcv_wnd * mss; |
843f4a55e
|
376 |
/* tcp_full_space because it is guaranteed to be the first packet */ |
ceef9ab6b
|
377 |
tcp_select_initial_window(sk_listener, full_space, |
843f4a55e
|
378 |
mss - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0), |
ed53d0ab7
|
379 380 |
&req->rsk_rcv_wnd, &req->rsk_window_clamp, |
843f4a55e
|
381 382 |
ireq->wscale_ok, &rcv_wscale, |
13d3b1ebe
|
383 |
rcv_wnd); |
843f4a55e
|
384 385 386 |
ireq->rcv_wscale = rcv_wscale; } EXPORT_SYMBOL(tcp_openreq_init_rwin); |
735d38311
|
387 388 |
static void tcp_ecn_openreq_child(struct tcp_sock *tp, const struct request_sock *req) |
bdf1ee5d3
|
389 390 391 |
{ tp->ecn_flags = inet_rsk(req)->ecn_ok ? TCP_ECN_OK : 0; } |
81164413a
|
392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 |
void tcp_ca_openreq_child(struct sock *sk, const struct dst_entry *dst) { struct inet_connection_sock *icsk = inet_csk(sk); u32 ca_key = dst_metric(dst, RTAX_CC_ALGO); bool ca_got_dst = false; if (ca_key != TCP_CA_UNSPEC) { const struct tcp_congestion_ops *ca; rcu_read_lock(); ca = tcp_ca_find_key(ca_key); if (likely(ca && try_module_get(ca->owner))) { icsk->icsk_ca_dst_locked = tcp_ca_dst_locked(dst); icsk->icsk_ca_ops = ca; ca_got_dst = true; } rcu_read_unlock(); } |
9f950415e
|
410 411 412 413 |
/* If no valid choice made yet, assign current system default ca. */ if (!ca_got_dst && (!icsk->icsk_ca_setsockopt || !try_module_get(icsk->icsk_ca_ops->owner))) |
81164413a
|
414 415 416 417 418 |
tcp_assign_congestion_control(sk); tcp_set_ca_state(sk, TCP_CA_Open); } EXPORT_SYMBOL_GPL(tcp_ca_openreq_child); |
60e2a7780
|
419 420 421 422 423 424 425 426 427 428 429 430 431 432 |
static void smc_check_reset_syn_req(struct tcp_sock *oldtp, struct request_sock *req, struct tcp_sock *newtp) { #if IS_ENABLED(CONFIG_SMC) struct inet_request_sock *ireq; if (static_branch_unlikely(&tcp_have_smc)) { ireq = inet_rsk(req); if (oldtp->syn_smc && !ireq->smc_ok) newtp->syn_smc = 0; } #endif } |
1da177e4c
|
433 434 435 436 437 438 |
/* This is not only more efficient than what we used to do, it eliminates * a lot of code duplication between IPv4/IPv6 SYN recv processing. -DaveM * * Actually, we could lots of memory writes here. tp of listening * socket contains all necessary default parameters. */ |
c28c6f045
|
439 440 441 |
struct sock *tcp_create_openreq_child(const struct sock *sk, struct request_sock *req, struct sk_buff *skb) |
1da177e4c
|
442 |
{ |
e56c57d0d
|
443 |
struct sock *newsk = inet_csk_clone_lock(sk, req, GFP_ATOMIC); |
242b1bbe5
|
444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 |
const struct inet_request_sock *ireq = inet_rsk(req); struct tcp_request_sock *treq = tcp_rsk(req); struct inet_connection_sock *newicsk; struct tcp_sock *oldtp, *newtp; if (!newsk) return NULL; newicsk = inet_csk(newsk); newtp = tcp_sk(newsk); oldtp = tcp_sk(sk); smc_check_reset_syn_req(oldtp, req, newtp); /* Now setup tcp_sock */ newtp->pred_flags = 0; newtp->rcv_wup = newtp->copied_seq = newtp->rcv_nxt = treq->rcv_isn + 1; newtp->segs_in = 1; newtp->snd_sml = newtp->snd_una = newtp->snd_nxt = newtp->snd_up = treq->snt_isn + 1; INIT_LIST_HEAD(&newtp->tsq_node); INIT_LIST_HEAD(&newtp->tsorted_sent_queue); tcp_init_wl(newtp, treq->rcv_isn); |
242b1bbe5
|
472 |
minmax_reset(&newtp->rtt_min, tcp_jiffies32, ~0U); |
242b1bbe5
|
473 |
newicsk->icsk_ack.lrcvtime = tcp_jiffies32; |
242b1bbe5
|
474 475 |
newtp->lsndtime = tcp_jiffies32; newsk->sk_txhash = treq->txhash; |
242b1bbe5
|
476 |
newtp->total_retrans = req->num_retrans; |
242b1bbe5
|
477 478 |
tcp_init_xmit_timers(newsk); newtp->write_seq = newtp->pushed_seq = treq->snt_isn + 1; |
242b1bbe5
|
479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 |
if (sock_flag(newsk, SOCK_KEEPOPEN)) inet_csk_reset_keepalive_timer(newsk, keepalive_time_when(newtp)); newtp->rx_opt.tstamp_ok = ireq->tstamp_ok; newtp->rx_opt.sack_ok = ireq->sack_ok; newtp->window_clamp = req->rsk_window_clamp; newtp->rcv_ssthresh = req->rsk_rcv_wnd; newtp->rcv_wnd = req->rsk_rcv_wnd; newtp->rx_opt.wscale_ok = ireq->wscale_ok; if (newtp->rx_opt.wscale_ok) { newtp->rx_opt.snd_wscale = ireq->snd_wscale; newtp->rx_opt.rcv_wscale = ireq->rcv_wscale; } else { newtp->rx_opt.snd_wscale = newtp->rx_opt.rcv_wscale = 0; newtp->window_clamp = min(newtp->window_clamp, 65535U); } newtp->snd_wnd = ntohs(tcp_hdr(skb)->window) << newtp->rx_opt.snd_wscale; newtp->max_window = newtp->snd_wnd; if (newtp->rx_opt.tstamp_ok) { newtp->rx_opt.ts_recent = req->ts_recent; |
cca9bab1b
|
501 |
newtp->rx_opt.ts_recent_stamp = ktime_get_seconds(); |
242b1bbe5
|
502 503 504 505 506 |
newtp->tcp_header_len = sizeof(struct tcphdr) + TCPOLEN_TSTAMP_ALIGNED; } else { newtp->rx_opt.ts_recent_stamp = 0; newtp->tcp_header_len = sizeof(struct tcphdr); } |
336c39a03
|
507 508 509 510 511 |
if (req->num_timeout) { newtp->undo_marker = treq->snt_isn; newtp->retrans_stamp = div_u64(treq->snt_synack, USEC_PER_SEC / TCP_TS_HZ); } |
242b1bbe5
|
512 |
newtp->tsoffset = treq->ts_off; |
cfb6eeb4c
|
513 |
#ifdef CONFIG_TCP_MD5SIG |
242b1bbe5
|
514 515 516 |
newtp->md5sig_info = NULL; /*XXX*/ if (newtp->af_specific->md5_lookup(sk, newsk)) newtp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED; |
cfb6eeb4c
|
517 |
#endif |
242b1bbe5
|
518 519 520 521 522 523 |
if (skb->len >= TCP_MSS_DEFAULT + newtp->tcp_header_len) newicsk->icsk_ack.last_seg_size = skb->len - newtp->tcp_header_len; newtp->rx_opt.mss_clamp = req->mss; tcp_ecn_openreq_child(newtp, req); newtp->fastopen_req = NULL; newtp->fastopen_rsk = NULL; |
242b1bbe5
|
524 525 |
__TCP_INC_STATS(sock_net(sk), TCP_MIB_PASSIVEOPENS); |
1da177e4c
|
526 527 |
return newsk; } |
4bc2f18ba
|
528 |
EXPORT_SYMBOL(tcp_create_openreq_child); |
1da177e4c
|
529 |
|
e905a9eda
|
530 |
/* |
8336886f7
|
531 532 533 534 535 536 |
* Process an incoming packet for SYN_RECV sockets represented as a * request_sock. Normally sk is the listener socket but for TFO it * points to the child socket. * * XXX (TFO) - The current impl contains a special check for ack * validation and inside tcp_v4_reqsk_send_ack(). Can we do better? |
4308fc58d
|
537 538 |
* * We don't need to initialize tmp_opt.sack_ok as we don't use the results |
1da177e4c
|
539 |
*/ |
5a5f3a8db
|
540 |
struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, |
60236fdd0
|
541 |
struct request_sock *req, |
e0f9759f5
|
542 |
bool fastopen, bool *req_stolen) |
1da177e4c
|
543 |
{ |
4957faade
|
544 |
struct tcp_options_received tmp_opt; |
4957faade
|
545 |
struct sock *child; |
aa8223c7b
|
546 |
const struct tcphdr *th = tcp_hdr(skb); |
714e85be3
|
547 |
__be32 flg = tcp_flag_word(th) & (TCP_FLAG_RST|TCP_FLAG_SYN|TCP_FLAG_ACK); |
a2a385d62
|
548 |
bool paws_reject = false; |
5e0724d02
|
549 |
bool own_req; |
1da177e4c
|
550 |
|
bb5b7c112
|
551 552 |
tmp_opt.saw_tstamp = 0; if (th->doff > (sizeof(struct tcphdr)>>2)) { |
eed29f17f
|
553 |
tcp_parse_options(sock_net(sk), skb, &tmp_opt, 0, NULL); |
1da177e4c
|
554 555 556 |
if (tmp_opt.saw_tstamp) { tmp_opt.ts_recent = req->ts_recent; |
95a22caee
|
557 558 |
if (tmp_opt.rcv_tsecr) tmp_opt.rcv_tsecr -= tcp_rsk(req)->ts_off; |
1da177e4c
|
559 560 561 562 |
/* We do not store true stamp, but it is not required, * it can be estimated (approximately) * from another data. */ |
cca9bab1b
|
563 |
tmp_opt.ts_recent_stamp = ktime_get_seconds() - ((TCP_TIMEOUT_INIT/HZ)<<req->num_timeout); |
c887e6d2d
|
564 |
paws_reject = tcp_paws_reject(&tmp_opt, th->rst); |
1da177e4c
|
565 566 567 568 |
} } /* Check for pure retransmitted SYN. */ |
2e6599cb8
|
569 |
if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn && |
1da177e4c
|
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 |
flg == TCP_FLAG_SYN && !paws_reject) { /* * RFC793 draws (Incorrectly! It was fixed in RFC1122) * this case on figure 6 and figure 8, but formal * protocol description says NOTHING. * To be more exact, it says that we should send ACK, * because this segment (at least, if it has no data) * is out of window. * * CONCLUSION: RFC793 (even with RFC1122) DOES NOT * describe SYN-RECV state. All the description * is wrong, we cannot believe to it and should * rely only on common sense and implementation * experience. * * Enforce "SYN-ACK" according to figure 8, figure 6 * of RFC793, fixed by RFC1122. |
8336886f7
|
588 589 590 |
* * Note that even if there is new data in the SYN packet * they will be thrown away too. |
cd75eff64
|
591 592 593 |
* * Reset timer after retransmitting SYNACK, similar to * the idea of fast retransmit in recovery. |
1da177e4c
|
594 |
*/ |
a9b2c06db
|
595 596 597 |
if (!tcp_oow_rate_limited(sock_net(sk), skb, LINUX_MIB_TCPACKSKIPPEDSYNRECV, &tcp_rsk(req)->last_oow_ack_time) && |
dd929c1b3
|
598 599 600 601 602 603 604 605 606 607 |
!inet_rtx_syn_ack(sk, req)) { unsigned long expires = jiffies; expires += min(TCP_TIMEOUT_INIT << req->num_timeout, TCP_RTO_MAX); if (!fastopen) mod_timer_pending(&req->rsk_timer, expires); else req->rsk_timer.expires = expires; } |
1da177e4c
|
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 |
return NULL; } /* Further reproduces section "SEGMENT ARRIVES" for state SYN-RECEIVED of RFC793. It is broken, however, it does not work only when SYNs are crossed. You would think that SYN crossing is impossible here, since we should have a SYN_SENT socket (from connect()) on our end, but this is not true if the crossed SYNs were sent to both ends by a malicious third party. We must defend against this, and to do that we first verify the ACK (as per RFC793, page 36) and reset if it is invalid. Is this a true full defense? To convince ourselves, let us consider a way in which the ACK test can still pass in this 'malicious crossed SYNs' case. Malicious sender sends identical SYNs (and thus identical sequence numbers) to both A and B: A: gets SYN, seq=7 B: gets SYN, seq=7 By our good fortune, both A and B select the same initial send sequence number of seven :-) A: sends SYN|ACK, seq=7, ack_seq=8 B: sends SYN|ACK, seq=7, ack_seq=8 So we are now A eating this SYN|ACK, ACK test passes. So does sequence test, SYN is truncated, and thus we consider it a bare ACK. |
ec0a19662
|
639 640 641 642 |
If icsk->icsk_accept_queue.rskq_defer_accept, we silently drop this bare ACK. Otherwise, we create an established connection. Both ends (listening sockets) accept the new incoming connection and try to talk to each other. 8-) |
1da177e4c
|
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 |
Note: This case is both harmless, and rare. Possibility is about the same as us discovering intelligent life on another plant tomorrow. But generally, we should (RFC lies!) to accept ACK from SYNACK both here and in tcp_rcv_state_process(). tcp_rcv_state_process() does not, hence, we do not too. Note that the case is absolutely generic: we cannot optimize anything here without violating protocol. All the checks must be made before attempt to create socket. */ /* RFC793 page 36: "If the connection is in any non-synchronized state ... * and the incoming segment acknowledges something not yet |
caa20d9ab
|
659 |
* sent (the segment carries an unacceptable ACK) ... |
1da177e4c
|
660 661 |
* a reset is sent." * |
8336886f7
|
662 663 664 665 |
* Invalid ACK: reset will be sent by listening socket. * Note that the ACK validity check for a Fast Open socket is done * elsewhere and is checked directly against the child socket rather * than req because user data may have been sent out. |
1da177e4c
|
666 |
*/ |
8336886f7
|
667 |
if ((flg & TCP_FLAG_ACK) && !fastopen && |
435cf559f
|
668 |
(TCP_SKB_CB(skb)->ack_seq != |
1a2c6181c
|
669 |
tcp_rsk(req)->snt_isn + 1)) |
1da177e4c
|
670 671 672 673 674 675 676 677 678 679 |
return sk; /* Also, it would be not so bad idea to check rcv_tsecr, which * is essentially ACK extension and too early or too late values * should cause reset in unsynchronized states. */ /* RFC793: "first check sequence number". */ if (paws_reject || !tcp_in_window(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq, |
ed53d0ab7
|
680 |
tcp_rsk(req)->rcv_nxt, tcp_rsk(req)->rcv_nxt + req->rsk_rcv_wnd)) { |
1da177e4c
|
681 |
/* Out of window: send ACK and drop. */ |
4ce7e93cb
|
682 683 684 685 |
if (!(flg & TCP_FLAG_RST) && !tcp_oow_rate_limited(sock_net(sk), skb, LINUX_MIB_TCPACKSKIPPEDSYNRECV, &tcp_rsk(req)->last_oow_ack_time)) |
6edafaaf6
|
686 |
req->rsk_ops->send_ack(sk, skb, req); |
1da177e4c
|
687 |
if (paws_reject) |
02a1d6e7a
|
688 |
__NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED); |
1da177e4c
|
689 690 691 692 |
return NULL; } /* In sequence, PAWS is OK. */ |
8336886f7
|
693 |
if (tmp_opt.saw_tstamp && !after(TCP_SKB_CB(skb)->seq, tcp_rsk(req)->rcv_nxt)) |
2aaab9a0c
|
694 |
req->ts_recent = tmp_opt.rcv_tsval; |
1da177e4c
|
695 |
|
2aaab9a0c
|
696 697 698 699 700 |
if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn) { /* Truncate SYN, it is out of window starting at tcp_rsk(req)->rcv_isn + 1. */ flg &= ~TCP_FLAG_SYN; } |
1da177e4c
|
701 |
|
2aaab9a0c
|
702 703 704 705 |
/* RFC793: "second check the RST bit" and * "fourth, check the SYN bit" */ if (flg & (TCP_FLAG_RST|TCP_FLAG_SYN)) { |
90bbcc608
|
706 |
__TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS); |
2aaab9a0c
|
707 708 |
goto embryonic_reset; } |
1da177e4c
|
709 |
|
2aaab9a0c
|
710 711 |
/* ACK sequence verified above, just make sure ACK is * set. If ACK not set, just silently drop the packet. |
8336886f7
|
712 713 714 |
* * XXX (TFO) - if we ever allow "data after SYN", the * following check needs to be removed. |
2aaab9a0c
|
715 716 717 |
*/ if (!(flg & TCP_FLAG_ACK)) return NULL; |
ec0a19662
|
718 |
|
8336886f7
|
719 720 721 722 723 |
/* For Fast Open no more processing is needed (sk is the * child socket). */ if (fastopen) return sk; |
d1b99ba41
|
724 |
/* While TCP_DEFER_ACCEPT is active, drop bare ACK. */ |
e6c022a4f
|
725 |
if (req->num_timeout < inet_csk(sk)->icsk_accept_queue.rskq_defer_accept && |
2aaab9a0c
|
726 727 |
TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) { inet_rsk(req)->acked = 1; |
02a1d6e7a
|
728 |
__NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPDEFERACCEPTDROP); |
2aaab9a0c
|
729 730 731 732 733 734 735 736 737 |
return NULL; } /* OK, ACK is valid, create big socket and * feed this segment to it. It will repeat all * the tests. THIS SEGMENT MUST MOVE SOCKET TO * ESTABLISHED STATE. If it will be dropped after * socket is created, wait for troubles. */ |
5e0724d02
|
738 739 |
child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb, req, NULL, req, &own_req); |
51456b291
|
740 |
if (!child) |
2aaab9a0c
|
741 |
goto listen_overflow; |
1da177e4c
|
742 |
|
6bcfd7f8c
|
743 |
sock_rps_save_rxhash(child, skb); |
0f1c28ae7
|
744 |
tcp_synack_rtt_meas(child, req); |
e0f9759f5
|
745 |
*req_stolen = !own_req; |
5e0724d02
|
746 |
return inet_csk_complete_hashdance(sk, child, req, own_req); |
1da177e4c
|
747 |
|
2aaab9a0c
|
748 |
listen_overflow: |
65c9410cf
|
749 |
if (!sock_net(sk)->ipv4.sysctl_tcp_abort_on_overflow) { |
2aaab9a0c
|
750 751 752 |
inet_rsk(req)->acked = 1; return NULL; } |
1da177e4c
|
753 |
|
2aaab9a0c
|
754 |
embryonic_reset: |
8336886f7
|
755 756 757 758 759 760 |
if (!(flg & TCP_FLAG_RST)) { /* Received a bad SYN pkt - for TFO We try not to reset * the local connection unless it's really necessary to * avoid becoming vulnerable to outside attack aiming at * resetting legit local connections. */ |
2aaab9a0c
|
761 |
req->rsk_ops->send_reset(sk, skb); |
8336886f7
|
762 763 764 765 766 |
} else if (fastopen) { /* received a valid RST pkt */ reqsk_fastopen_remove(sk, req, true); tcp_reset(sk); } if (!fastopen) { |
52452c542
|
767 |
inet_csk_reqsk_queue_drop(sk, req); |
02a1d6e7a
|
768 |
__NET_INC_STATS(sock_net(sk), LINUX_MIB_EMBRYONICRSTS); |
8336886f7
|
769 |
} |
2aaab9a0c
|
770 |
return NULL; |
1da177e4c
|
771 |
} |
4bc2f18ba
|
772 |
EXPORT_SYMBOL(tcp_check_req); |
1da177e4c
|
773 774 775 776 777 |
/* * Queue segment on the new socket if the new socket is active, * otherwise we just shortcircuit this and continue with * the new socket. |
8336886f7
|
778 779 780 781 782 783 |
* * For the vast majority of cases child->sk_state will be TCP_SYN_RECV * when entering. But other states are possible due to a race condition * where after __inet_lookup_established() fails but before the listener * locked is obtained, other packets cause the same connection to * be created. |
1da177e4c
|
784 785 786 787 788 789 790 |
*/ int tcp_child_process(struct sock *parent, struct sock *child, struct sk_buff *skb) { int ret = 0; int state = child->sk_state; |
e5907459c
|
791 792 |
/* record NAPI ID of child */ sk_mark_napi_id(child, skb); |
a44d6eacd
|
793 |
tcp_segs_in(tcp_sk(child), skb); |
1da177e4c
|
794 |
if (!sock_owned_by_user(child)) { |
72ab4a86f
|
795 |
ret = tcp_rcv_state_process(child, skb); |
1da177e4c
|
796 797 |
/* Wakeup parent, send SIGIO */ if (state == TCP_SYN_RECV && child->sk_state != state) |
676d23690
|
798 |
parent->sk_data_ready(parent); |
1da177e4c
|
799 800 801 802 803 |
} else { /* Alas, it is possible again, because we do lookup * in main socket hash table and lock on listening * socket does not protect us more. */ |
a3a858ff1
|
804 |
__sk_add_backlog(child, skb); |
1da177e4c
|
805 806 807 808 809 810 |
} bh_unlock_sock(child); sock_put(child); return ret; } |
1da177e4c
|
811 |
EXPORT_SYMBOL(tcp_child_process); |