tcp_output.c 81.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 * 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).
 *
8
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
 *		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>
 */

/*
 * Changes:	Pedro Roque	:	Retransmit queue handled by TCP.
 *				:	Fragmentation on mtu decrease
 *				:	Segment collapse on retransmit
 *				:	AF independence
 *
 *		Linus Torvalds	:	send_delayed_ack
 *		David S. Miller	:	Charge memory using the right skb
 *					during syn/ack processing.
 *		David S. Miller :	Output engine completely rewritten.
 *		Andrea Arcangeli:	SYNACK carry ts_recent in tsecr.
 *		Cacophonix Gaul :	draft-minshall-nagle-01
 *		J Hadi Salim	:	ECN support
 *
 */

#include <net/tcp.h>

#include <linux/compiler.h>
#include <linux/module.h>

/* People can turn this off for buggy TCP's found in printers etc. */
43
int sysctl_tcp_retrans_collapse __read_mostly = 1;
L
Linus Torvalds 已提交
44

45
/* People can turn this on to work with those rare, broken TCPs that
46 47
 * interpret the window field as a signed quantity.
 */
48
int sysctl_tcp_workaround_signed_windows __read_mostly = 0;
49

L
Linus Torvalds 已提交
50 51 52 53
/* This limits the percentage of the congestion window which we
 * will allow a single TSO frame to consume.  Building TSO frames
 * which are too large can cause TCP streams to be bursty.
 */
54
int sysctl_tcp_tso_win_divisor __read_mostly = 3;
L
Linus Torvalds 已提交
55

56 57
int sysctl_tcp_mtu_probing __read_mostly = 0;
int sysctl_tcp_base_mss __read_mostly = 512;
J
John Heffner 已提交
58

59
/* By default, RFC2861 behavior.  */
60
int sysctl_tcp_slow_start_after_idle __read_mostly = 1;
61

62
int sysctl_tcp_cookie_size __read_mostly = 0; /* TCP_COOKIE_MAX */
63
EXPORT_SYMBOL_GPL(sysctl_tcp_cookie_size);
64 65


66
/* Account for new data that has been sent to the network. */
67
static void tcp_event_new_data_sent(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
68
{
69
	struct tcp_sock *tp = tcp_sk(sk);
70
	unsigned int prior_packets = tp->packets_out;
71

72
	tcp_advance_send_head(sk, skb);
L
Linus Torvalds 已提交
73
	tp->snd_nxt = TCP_SKB_CB(skb)->end_seq;
74 75 76 77

	/* Don't override Nagle indefinately with F-RTO */
	if (tp->frto_counter == 2)
		tp->frto_counter = 3;
78 79 80 81 82

	tp->packets_out += tcp_skb_pcount(skb);
	if (!prior_packets)
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
					  inet_csk(sk)->icsk_rto, TCP_RTO_MAX);
L
Linus Torvalds 已提交
83 84 85 86 87 88 89 90
}

/* SND.NXT, if window was not shrunk.
 * If window has been shrunk, what should we make? It is not clear at all.
 * Using SND.UNA we will fail to open window, SND.NXT is out of window. :-(
 * Anything in between SND.UNA...SND.UNA+SND.WND also can be already
 * invalid. OK, let's make this for now:
 */
91
static inline __u32 tcp_acceptable_seq(struct sock *sk)
L
Linus Torvalds 已提交
92
{
93 94
	struct tcp_sock *tp = tcp_sk(sk);

95
	if (!before(tcp_wnd_end(tp), tp->snd_nxt))
L
Linus Torvalds 已提交
96 97
		return tp->snd_nxt;
	else
98
		return tcp_wnd_end(tp);
L
Linus Torvalds 已提交
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
}

/* Calculate mss to advertise in SYN segment.
 * RFC1122, RFC1063, draft-ietf-tcpimpl-pmtud-01 state that:
 *
 * 1. It is independent of path mtu.
 * 2. Ideally, it is maximal possible segment size i.e. 65535-40.
 * 3. For IPv4 it is reasonable to calculate it from maximal MTU of
 *    attached devices, because some buggy hosts are confused by
 *    large MSS.
 * 4. We do not make 3, we advertise MSS, calculated from first
 *    hop device mtu, but allow to raise it to ip_rt_min_advmss.
 *    This may be overridden via information stored in routing table.
 * 5. Value 65535 for MSS is valid in IPv6 and means "as large as possible,
 *    probably even Jumbo".
 */
static __u16 tcp_advertise_mss(struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct dst_entry *dst = __sk_dst_get(sk);
	int mss = tp->advmss;

	if (dst && dst_metric(dst, RTAX_ADVMSS) < mss) {
		mss = dst_metric(dst, RTAX_ADVMSS);
		tp->advmss = mss;
	}

	return (__u16)mss;
}

/* RFC2861. Reset CWND after idle period longer RTO to "restart window".
 * This is the first part of cwnd validation mechanism. */
131
static void tcp_cwnd_restart(struct sock *sk, struct dst_entry *dst)
L
Linus Torvalds 已提交
132
{
133
	struct tcp_sock *tp = tcp_sk(sk);
L
Linus Torvalds 已提交
134 135 136 137
	s32 delta = tcp_time_stamp - tp->lsndtime;
	u32 restart_cwnd = tcp_init_cwnd(tp, dst);
	u32 cwnd = tp->snd_cwnd;

138
	tcp_ca_event(sk, CA_EVENT_CWND_RESTART);
L
Linus Torvalds 已提交
139

140
	tp->snd_ssthresh = tcp_current_ssthresh(sk);
L
Linus Torvalds 已提交
141 142
	restart_cwnd = min(restart_cwnd, cwnd);

143
	while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd)
L
Linus Torvalds 已提交
144 145 146 147 148 149
		cwnd >>= 1;
	tp->snd_cwnd = max(cwnd, restart_cwnd);
	tp->snd_cwnd_stamp = tcp_time_stamp;
	tp->snd_cwnd_used = 0;
}

150
/* Congestion state accounting after a packet has been sent. */
S
Stephen Hemminger 已提交
151 152
static void tcp_event_data_sent(struct tcp_sock *tp,
				struct sk_buff *skb, struct sock *sk)
L
Linus Torvalds 已提交
153
{
154 155
	struct inet_connection_sock *icsk = inet_csk(sk);
	const u32 now = tcp_time_stamp;
L
Linus Torvalds 已提交
156

157 158
	if (sysctl_tcp_slow_start_after_idle &&
	    (!tp->packets_out && (s32)(now - tp->lsndtime) > icsk->icsk_rto))
159
		tcp_cwnd_restart(sk, __sk_dst_get(sk));
L
Linus Torvalds 已提交
160 161 162 163 164 165

	tp->lsndtime = now;

	/* If it is a reply for ato after last received
	 * packet, enter pingpong mode.
	 */
166 167
	if ((u32)(now - icsk->icsk_ack.lrcvtime) < icsk->icsk_ack.ato)
		icsk->icsk_ack.pingpong = 1;
L
Linus Torvalds 已提交
168 169
}

170
/* Account for an ACK we sent. */
S
Stephen Hemminger 已提交
171
static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts)
L
Linus Torvalds 已提交
172
{
173 174
	tcp_dec_quickack_mode(sk, pkts);
	inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
L
Linus Torvalds 已提交
175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199
}

/* Determine a window scaling and initial window to offer.
 * Based on the assumption that the given amount of space
 * will be offered. Store the results in the tp structure.
 * NOTE: for smooth operation initial space offering should
 * be a multiple of mss if possible. We assume here that mss >= 1.
 * This MUST be enforced by all callers.
 */
void tcp_select_initial_window(int __space, __u32 mss,
			       __u32 *rcv_wnd, __u32 *window_clamp,
			       int wscale_ok, __u8 *rcv_wscale)
{
	unsigned int space = (__space < 0 ? 0 : __space);

	/* If no clamp set the clamp to the max possible scaled window */
	if (*window_clamp == 0)
		(*window_clamp) = (65535 << 14);
	space = min(*window_clamp, space);

	/* Quantize space offering to a multiple of mss if possible. */
	if (space > mss)
		space = (space / mss) * mss;

	/* NOTE: offering an initial window larger than 32767
200 201 202 203 204 205
	 * will break some buggy TCP stacks. If the admin tells us
	 * it is likely we could be speaking with such a buggy stack
	 * we will truncate our initial window offering to 32K-1
	 * unless the remote has sent us a window scaling option,
	 * which we interpret as a sign the remote TCP is not
	 * misinterpreting the window field as a signed quantity.
L
Linus Torvalds 已提交
206
	 */
207 208 209 210 211
	if (sysctl_tcp_workaround_signed_windows)
		(*rcv_wnd) = min(space, MAX_TCP_WINDOW);
	else
		(*rcv_wnd) = space;

L
Linus Torvalds 已提交
212 213 214
	(*rcv_wscale) = 0;
	if (wscale_ok) {
		/* Set window scaling on max possible window
215
		 * See RFC1323 for an explanation of the limit to 14
L
Linus Torvalds 已提交
216 217
		 */
		space = max_t(u32, sysctl_tcp_rmem[2], sysctl_rmem_max);
218
		space = min_t(u32, space, *window_clamp);
L
Linus Torvalds 已提交
219 220 221 222 223 224 225
		while (space > 65535 && (*rcv_wscale) < 14) {
			space >>= 1;
			(*rcv_wscale)++;
		}
	}

	/* Set initial window to value enough for senders,
226
	 * following RFC2414. Senders, not following this RFC,
L
Linus Torvalds 已提交
227 228
	 * will be satisfied with 2.
	 */
229
	if (mss > (1 << *rcv_wscale)) {
230
		int init_cwnd = 4;
231
		if (mss > 1460 * 3)
L
Linus Torvalds 已提交
232
			init_cwnd = 2;
233 234
		else if (mss > 1460)
			init_cwnd = 3;
235 236
		if (*rcv_wnd > init_cwnd * mss)
			*rcv_wnd = init_cwnd * mss;
L
Linus Torvalds 已提交
237 238 239 240 241 242 243 244 245 246 247
	}

	/* Set the clamp no higher than max representable value */
	(*window_clamp) = min(65535U << (*rcv_wscale), *window_clamp);
}

/* Chose a new window to advertise, update state in tcp_sock for the
 * socket, and return result with RFC1323 scaling applied.  The return
 * value can be stuffed directly into th->window for an outgoing
 * frame.
 */
S
Stephen Hemminger 已提交
248
static u16 tcp_select_window(struct sock *sk)
L
Linus Torvalds 已提交
249 250 251 252 253 254
{
	struct tcp_sock *tp = tcp_sk(sk);
	u32 cur_win = tcp_receive_window(tp);
	u32 new_win = __tcp_select_window(sk);

	/* Never shrink the offered window */
S
Stephen Hemminger 已提交
255
	if (new_win < cur_win) {
L
Linus Torvalds 已提交
256 257 258 259 260 261 262
		/* Danger Will Robinson!
		 * Don't update rcv_wup/rcv_wnd here or else
		 * we will not be able to advertise a zero
		 * window in time.  --DaveM
		 *
		 * Relax Will Robinson.
		 */
263
		new_win = ALIGN(cur_win, 1 << tp->rx_opt.rcv_wscale);
L
Linus Torvalds 已提交
264 265 266 267 268 269 270
	}
	tp->rcv_wnd = new_win;
	tp->rcv_wup = tp->rcv_nxt;

	/* Make sure we do not exceed the maximum possible
	 * scaled window.
	 */
271
	if (!tp->rx_opt.rcv_wscale && sysctl_tcp_workaround_signed_windows)
L
Linus Torvalds 已提交
272 273 274 275 276 277 278 279 280 281 282 283 284 285
		new_win = min(new_win, MAX_TCP_WINDOW);
	else
		new_win = min(new_win, (65535U << tp->rx_opt.rcv_wscale));

	/* RFC1323 scaling applied */
	new_win >>= tp->rx_opt.rcv_wscale;

	/* If we advertise zero window, disable fast path. */
	if (new_win == 0)
		tp->pred_flags = 0;

	return new_win;
}

286
/* Packet ECN state for a SYN-ACK */
287
static inline void TCP_ECN_send_synack(struct tcp_sock *tp, struct sk_buff *skb)
288 289
{
	TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_CWR;
290
	if (!(tp->ecn_flags & TCP_ECN_OK))
291 292 293
		TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_ECE;
}

294
/* Packet ECN state for a SYN.  */
295 296 297 298 299
static inline void TCP_ECN_send_syn(struct sock *sk, struct sk_buff *skb)
{
	struct tcp_sock *tp = tcp_sk(sk);

	tp->ecn_flags = 0;
300
	if (sysctl_tcp_ecn == 1) {
301
		TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_ECE | TCPCB_FLAG_CWR;
302 303 304 305 306 307 308 309 310 311 312
		tp->ecn_flags = TCP_ECN_OK;
	}
}

static __inline__ void
TCP_ECN_make_synack(struct request_sock *req, struct tcphdr *th)
{
	if (inet_rsk(req)->ecn_ok)
		th->ece = 1;
}

313 314 315
/* Set up ECN state for a packet on a ESTABLISHED socket that is about to
 * be sent.
 */
316 317 318 319 320 321 322 323 324 325
static inline void TCP_ECN_send(struct sock *sk, struct sk_buff *skb,
				int tcp_header_len)
{
	struct tcp_sock *tp = tcp_sk(sk);

	if (tp->ecn_flags & TCP_ECN_OK) {
		/* Not-retransmitted data segment: set ECT and inject CWR. */
		if (skb->len != tcp_header_len &&
		    !before(TCP_SKB_CB(skb)->seq, tp->snd_nxt)) {
			INET_ECN_xmit(sk);
326
			if (tp->ecn_flags & TCP_ECN_QUEUE_CWR) {
327 328 329 330 331 332 333 334 335 336 337 338 339
				tp->ecn_flags &= ~TCP_ECN_QUEUE_CWR;
				tcp_hdr(skb)->cwr = 1;
				skb_shinfo(skb)->gso_type |= SKB_GSO_TCP_ECN;
			}
		} else {
			/* ACK or retransmitted segment: clear ECT|CE */
			INET_ECN_dontxmit(sk);
		}
		if (tp->ecn_flags & TCP_ECN_DEMAND_CWR)
			tcp_hdr(skb)->ece = 1;
	}
}

340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
/* Constructs common control bits of non-data skb. If SYN/FIN is present,
 * auto increment end seqno.
 */
static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags)
{
	skb->csum = 0;

	TCP_SKB_CB(skb)->flags = flags;
	TCP_SKB_CB(skb)->sacked = 0;

	skb_shinfo(skb)->gso_segs = 1;
	skb_shinfo(skb)->gso_size = 0;
	skb_shinfo(skb)->gso_type = 0;

	TCP_SKB_CB(skb)->seq = seq;
	if (flags & (TCPCB_FLAG_SYN | TCPCB_FLAG_FIN))
		seq++;
	TCP_SKB_CB(skb)->end_seq = seq;
}

I
Ilpo Järvinen 已提交
360 361 362 363 364
static inline int tcp_urg_mode(const struct tcp_sock *tp)
{
	return tp->snd_una != tp->snd_up;
}

A
Adam Langley 已提交
365 366 367
#define OPTION_SACK_ADVERTISE	(1 << 0)
#define OPTION_TS		(1 << 1)
#define OPTION_MD5		(1 << 2)
368
#define OPTION_WSCALE		(1 << 3)
369
#define OPTION_COOKIE_EXTENSION	(1 << 4)
A
Adam Langley 已提交
370 371 372 373 374

struct tcp_out_options {
	u8 options;		/* bit field of OPTION_* */
	u8 ws;			/* window scale, 0 to disable */
	u8 num_sack_blocks;	/* number of SACK blocks to include */
375
	u8 hash_size;		/* bytes in hash_location */
A
Adam Langley 已提交
376 377
	u16 mss;		/* 0 to disable */
	__u32 tsval, tsecr;	/* need to include OPTION_TS */
378
	__u8 *hash_location;	/* temporary pointer, overloaded */
A
Adam Langley 已提交
379 380
};

381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
/* The sysctl int routines are generic, so check consistency here.
 */
static u8 tcp_cookie_size_check(u8 desired)
{
	if (desired > 0) {
		/* previously specified */
		return desired;
	}
	if (sysctl_tcp_cookie_size <= 0) {
		/* no default specified */
		return 0;
	}
	if (sysctl_tcp_cookie_size <= TCP_COOKIE_MIN) {
		/* value too small, specify minimum */
		return TCP_COOKIE_MIN;
	}
	if (sysctl_tcp_cookie_size >= TCP_COOKIE_MAX) {
		/* value too large, specify maximum */
		return TCP_COOKIE_MAX;
	}
	if (0x1 & sysctl_tcp_cookie_size) {
		/* 8-bit multiple, illegal, fix it */
		return (u8)(sysctl_tcp_cookie_size + 0x1);
	}
	return (u8)sysctl_tcp_cookie_size;
}

408 409 410
/* Write previously computed TCP options to the packet.
 *
 * Beware: Something in the Internet is very sensitive to the ordering of
411 412 413 414 415 416 417 418 419 420
 * TCP options, we learned this through the hard way, so be careful here.
 * Luckily we can at least blame others for their non-compliance but from
 * inter-operatibility perspective it seems that we're somewhat stuck with
 * the ordering which we have been using if we want to keep working with
 * those broken things (not that it currently hurts anybody as there isn't
 * particular reason why the ordering would need to be changed).
 *
 * At least SACK_PERM as the first option is known to lead to a disaster
 * (but it may well be that other scenarios fail similarly).
 */
A
Adam Langley 已提交
421
static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp,
422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
			      struct tcp_out_options *opts)
{
	u8 options = opts->options;	/* mungable copy */

	/* Having both authentication and cookies for security is redundant,
	 * and there's certainly not enough room.  Instead, the cookie-less
	 * extension variant is proposed.
	 *
	 * Consider the pessimal case with authentication.  The options
	 * could look like:
	 *   COOKIE|MD5(20) + MSS(4) + SACK|TS(12) + WSCALE(4) == 40
	 */
	if (unlikely(OPTION_MD5 & options)) {
		if (unlikely(OPTION_COOKIE_EXTENSION & options)) {
			*ptr++ = htonl((TCPOPT_COOKIE << 24) |
				       (TCPOLEN_COOKIE_BASE << 16) |
				       (TCPOPT_MD5SIG << 8) |
				       TCPOLEN_MD5SIG);
		} else {
			*ptr++ = htonl((TCPOPT_NOP << 24) |
				       (TCPOPT_NOP << 16) |
				       (TCPOPT_MD5SIG << 8) |
				       TCPOLEN_MD5SIG);
		}
		options &= ~OPTION_COOKIE_EXTENSION;
		/* overload cookie hash location */
		opts->hash_location = (__u8 *)ptr;
A
Adam Langley 已提交
449
		ptr += 4;
S
Stephen Hemminger 已提交
450
	}
A
Adam Langley 已提交
451

452 453 454 455 456 457
	if (unlikely(opts->mss)) {
		*ptr++ = htonl((TCPOPT_MSS << 24) |
			       (TCPOLEN_MSS << 16) |
			       opts->mss);
	}

458 459
	if (likely(OPTION_TS & options)) {
		if (unlikely(OPTION_SACK_ADVERTISE & options)) {
A
Adam Langley 已提交
460 461 462 463
			*ptr++ = htonl((TCPOPT_SACK_PERM << 24) |
				       (TCPOLEN_SACK_PERM << 16) |
				       (TCPOPT_TIMESTAMP << 8) |
				       TCPOLEN_TIMESTAMP);
464
			options &= ~OPTION_SACK_ADVERTISE;
A
Adam Langley 已提交
465 466 467 468 469 470 471 472 473 474
		} else {
			*ptr++ = htonl((TCPOPT_NOP << 24) |
				       (TCPOPT_NOP << 16) |
				       (TCPOPT_TIMESTAMP << 8) |
				       TCPOLEN_TIMESTAMP);
		}
		*ptr++ = htonl(opts->tsval);
		*ptr++ = htonl(opts->tsecr);
	}

475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513
	/* Specification requires after timestamp, so do it now.
	 *
	 * Consider the pessimal case without authentication.  The options
	 * could look like:
	 *   MSS(4) + SACK|TS(12) + COOKIE(20) + WSCALE(4) == 40
	 */
	if (unlikely(OPTION_COOKIE_EXTENSION & options)) {
		__u8 *cookie_copy = opts->hash_location;
		u8 cookie_size = opts->hash_size;

		/* 8-bit multiple handled in tcp_cookie_size_check() above,
		 * and elsewhere.
		 */
		if (0x2 & cookie_size) {
			__u8 *p = (__u8 *)ptr;

			/* 16-bit multiple */
			*p++ = TCPOPT_COOKIE;
			*p++ = TCPOLEN_COOKIE_BASE + cookie_size;
			*p++ = *cookie_copy++;
			*p++ = *cookie_copy++;
			ptr++;
			cookie_size -= 2;
		} else {
			/* 32-bit multiple */
			*ptr++ = htonl(((TCPOPT_NOP << 24) |
					(TCPOPT_NOP << 16) |
					(TCPOPT_COOKIE << 8) |
					TCPOLEN_COOKIE_BASE) +
				       cookie_size);
		}

		if (cookie_size > 0) {
			memcpy(ptr, cookie_copy, cookie_size);
			ptr += (cookie_size / 4);
		}
	}

	if (unlikely(OPTION_SACK_ADVERTISE & options)) {
A
Adam Langley 已提交
514 515 516 517 518 519
		*ptr++ = htonl((TCPOPT_NOP << 24) |
			       (TCPOPT_NOP << 16) |
			       (TCPOPT_SACK_PERM << 8) |
			       TCPOLEN_SACK_PERM);
	}

520
	if (unlikely(OPTION_WSCALE & options)) {
A
Adam Langley 已提交
521 522 523 524 525 526 527 528 529
		*ptr++ = htonl((TCPOPT_NOP << 24) |
			       (TCPOPT_WINDOW << 16) |
			       (TCPOLEN_WINDOW << 8) |
			       opts->ws);
	}

	if (unlikely(opts->num_sack_blocks)) {
		struct tcp_sack_block *sp = tp->rx_opt.dsack ?
			tp->duplicate_sack : tp->selective_acks;
S
Stephen Hemminger 已提交
530 531 532 533 534
		int this_sack;

		*ptr++ = htonl((TCPOPT_NOP  << 24) |
			       (TCPOPT_NOP  << 16) |
			       (TCPOPT_SACK <<  8) |
A
Adam Langley 已提交
535
			       (TCPOLEN_SACK_BASE + (opts->num_sack_blocks *
S
Stephen Hemminger 已提交
536
						     TCPOLEN_SACK_PERBLOCK)));
S
Stephen Hemminger 已提交
537

A
Adam Langley 已提交
538 539
		for (this_sack = 0; this_sack < opts->num_sack_blocks;
		     ++this_sack) {
S
Stephen Hemminger 已提交
540 541 542
			*ptr++ = htonl(sp[this_sack].start_seq);
			*ptr++ = htonl(sp[this_sack].end_seq);
		}
S
Stephen Hemminger 已提交
543

544
		tp->rx_opt.dsack = 0;
S
Stephen Hemminger 已提交
545
	}
A
Adam Langley 已提交
546 547
}

548 549 550
/* Compute TCP options for SYN packets. This is not the final
 * network wire format yet.
 */
A
Adam Langley 已提交
551 552 553 554
static unsigned tcp_syn_options(struct sock *sk, struct sk_buff *skb,
				struct tcp_out_options *opts,
				struct tcp_md5sig_key **md5) {
	struct tcp_sock *tp = tcp_sk(sk);
555
	struct tcp_cookie_values *cvp = tp->cookie_values;
556
	struct dst_entry *dst = __sk_dst_get(sk);
557 558 559 560
	unsigned remaining = MAX_TCP_OPTION_SPACE;
	u8 cookie_size = (!tp->rx_opt.cookie_out_never && cvp != NULL) ?
			 tcp_cookie_size_check(cvp->cookie_desired) :
			 0;
A
Adam Langley 已提交
561

562
#ifdef CONFIG_TCP_MD5SIG
A
Adam Langley 已提交
563 564 565
	*md5 = tp->af_specific->md5_lookup(sk, sk);
	if (*md5) {
		opts->options |= OPTION_MD5;
566
		remaining -= TCPOLEN_MD5SIG_ALIGNED;
567
	}
A
Adam Langley 已提交
568 569
#else
	*md5 = NULL;
570
#endif
A
Adam Langley 已提交
571 572 573 574 575 576 577 578 579 580 581

	/* We always get an MSS option.  The option bytes which will be seen in
	 * normal data packets should timestamps be used, must be in the MSS
	 * advertised.  But we subtract them from tp->mss_cache so that
	 * calculations in tcp_sendmsg are simpler etc.  So account for this
	 * fact here if necessary.  If we don't do this correctly, as a
	 * receiver we won't recognize data packets as being full sized when we
	 * should, and thus we won't abide by the delayed ACK rules correctly.
	 * SACKs don't matter, we never delay an ACK when we have any of those
	 * going out.  */
	opts->mss = tcp_advertise_mss(sk);
582
	remaining -= TCPOLEN_MSS_ALIGNED;
A
Adam Langley 已提交
583

584 585 586
	if (likely(sysctl_tcp_timestamps &&
		   !dst_feature(dst, RTAX_FEATURE_NO_TSTAMP) &&
		   *md5 == NULL)) {
A
Adam Langley 已提交
587 588 589
		opts->options |= OPTION_TS;
		opts->tsval = TCP_SKB_CB(skb)->when;
		opts->tsecr = tp->rx_opt.ts_recent;
590
		remaining -= TCPOLEN_TSTAMP_ALIGNED;
A
Adam Langley 已提交
591
	}
592 593
	if (likely(sysctl_tcp_window_scaling &&
		   !dst_feature(dst, RTAX_FEATURE_NO_WSCALE))) {
A
Adam Langley 已提交
594
		opts->ws = tp->rx_opt.rcv_wscale;
595
		opts->options |= OPTION_WSCALE;
596
		remaining -= TCPOLEN_WSCALE_ALIGNED;
A
Adam Langley 已提交
597
	}
598 599
	if (likely(sysctl_tcp_sack &&
		   !dst_feature(dst, RTAX_FEATURE_NO_SACK))) {
A
Adam Langley 已提交
600
		opts->options |= OPTION_SACK_ADVERTISE;
601
		if (unlikely(!(OPTION_TS & opts->options)))
602
			remaining -= TCPOLEN_SACKPERM_ALIGNED;
A
Adam Langley 已提交
603 604
	}

605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
	/* Note that timestamps are required by the specification.
	 *
	 * Odd numbers of bytes are prohibited by the specification, ensuring
	 * that the cookie is 16-bit aligned, and the resulting cookie pair is
	 * 32-bit aligned.
	 */
	if (*md5 == NULL &&
	    (OPTION_TS & opts->options) &&
	    cookie_size > 0) {
		int need = TCPOLEN_COOKIE_BASE + cookie_size;

		if (0x2 & need) {
			/* 32-bit multiple */
			need += 2; /* NOPs */

			if (need > remaining) {
				/* try shrinking cookie to fit */
				cookie_size -= 2;
				need -= 4;
			}
		}
		while (need > remaining && TCP_COOKIE_MIN <= cookie_size) {
			cookie_size -= 4;
			need -= 4;
		}
		if (TCP_COOKIE_MIN <= cookie_size) {
			opts->options |= OPTION_COOKIE_EXTENSION;
			opts->hash_location = (__u8 *)&cvp->cookie_pair[0];
			opts->hash_size = cookie_size;

			/* Remember for future incarnations. */
			cvp->cookie_desired = cookie_size;

			if (cvp->cookie_desired != cvp->cookie_pair_size) {
				/* Currently use random bytes as a nonce,
				 * assuming these are completely unpredictable
				 * by hostile users of the same system.
				 */
				get_random_bytes(&cvp->cookie_pair[0],
						 cookie_size);
				cvp->cookie_pair_size = cookie_size;
			}

			remaining -= need;
		}
	}
	return MAX_TCP_OPTION_SPACE - remaining;
S
Stephen Hemminger 已提交
652 653
}

654
/* Set up TCP options for SYN-ACKs. */
A
Adam Langley 已提交
655 656 657 658
static unsigned tcp_synack_options(struct sock *sk,
				   struct request_sock *req,
				   unsigned mss, struct sk_buff *skb,
				   struct tcp_out_options *opts,
659 660 661
				   struct tcp_md5sig_key **md5,
				   struct tcp_extend_values *xvp)
{
A
Adam Langley 已提交
662
	struct inet_request_sock *ireq = inet_rsk(req);
663 664 665 666 667
	unsigned remaining = MAX_TCP_OPTION_SPACE;
	u8 cookie_plus = (xvp != NULL && !xvp->cookie_out_never) ?
			 xvp->cookie_plus :
			 0;
	bool doing_ts = ireq->tstamp_ok;
A
Adam Langley 已提交
668

669
#ifdef CONFIG_TCP_MD5SIG
A
Adam Langley 已提交
670 671 672
	*md5 = tcp_rsk(req)->af_specific->md5_lookup(sk, req);
	if (*md5) {
		opts->options |= OPTION_MD5;
673 674 675 676 677 678 679 680
		remaining -= TCPOLEN_MD5SIG_ALIGNED;

		/* We can't fit any SACK blocks in a packet with MD5 + TS
		 * options. There was discussion about disabling SACK
		 * rather than TS in order to fit in better with old,
		 * buggy kernels, but that was deemed to be unnecessary.
		 */
		doing_ts &= !ireq->sack_ok;
681
	}
A
Adam Langley 已提交
682 683
#else
	*md5 = NULL;
684
#endif
A
Adam Langley 已提交
685

686
	/* We always send an MSS option. */
A
Adam Langley 已提交
687
	opts->mss = mss;
688
	remaining -= TCPOLEN_MSS_ALIGNED;
A
Adam Langley 已提交
689 690 691

	if (likely(ireq->wscale_ok)) {
		opts->ws = ireq->rcv_wscale;
692
		opts->options |= OPTION_WSCALE;
693
		remaining -= TCPOLEN_WSCALE_ALIGNED;
A
Adam Langley 已提交
694 695 696 697 698
	}
	if (likely(doing_ts)) {
		opts->options |= OPTION_TS;
		opts->tsval = TCP_SKB_CB(skb)->when;
		opts->tsecr = req->ts_recent;
699
		remaining -= TCPOLEN_TSTAMP_ALIGNED;
A
Adam Langley 已提交
700 701 702 703
	}
	if (likely(ireq->sack_ok)) {
		opts->options |= OPTION_SACK_ADVERTISE;
		if (unlikely(!doing_ts))
704
			remaining -= TCPOLEN_SACKPERM_ALIGNED;
A
Adam Langley 已提交
705 706
	}

707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729
	/* Similar rationale to tcp_syn_options() applies here, too.
	 * If the <SYN> options fit, the same options should fit now!
	 */
	if (*md5 == NULL &&
	    doing_ts &&
	    cookie_plus > TCPOLEN_COOKIE_BASE) {
		int need = cookie_plus; /* has TCPOLEN_COOKIE_BASE */

		if (0x2 & need) {
			/* 32-bit multiple */
			need += 2; /* NOPs */
		}
		if (need <= remaining) {
			opts->options |= OPTION_COOKIE_EXTENSION;
			opts->hash_size = cookie_plus - TCPOLEN_COOKIE_BASE;
			remaining -= need;
		} else {
			/* There's no error return, so flag it. */
			xvp->cookie_out_never = 1; /* true */
			opts->hash_size = 0;
		}
	}
	return MAX_TCP_OPTION_SPACE - remaining;
A
Adam Langley 已提交
730 731
}

732 733 734
/* Compute TCP options for ESTABLISHED sockets. This is not the
 * final wire format yet.
 */
A
Adam Langley 已提交
735 736 737 738 739 740
static unsigned tcp_established_options(struct sock *sk, struct sk_buff *skb,
					struct tcp_out_options *opts,
					struct tcp_md5sig_key **md5) {
	struct tcp_skb_cb *tcb = skb ? TCP_SKB_CB(skb) : NULL;
	struct tcp_sock *tp = tcp_sk(sk);
	unsigned size = 0;
741
	unsigned int eff_sacks;
A
Adam Langley 已提交
742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759

#ifdef CONFIG_TCP_MD5SIG
	*md5 = tp->af_specific->md5_lookup(sk, sk);
	if (unlikely(*md5)) {
		opts->options |= OPTION_MD5;
		size += TCPOLEN_MD5SIG_ALIGNED;
	}
#else
	*md5 = NULL;
#endif

	if (likely(tp->rx_opt.tstamp_ok)) {
		opts->options |= OPTION_TS;
		opts->tsval = tcb ? tcb->when : 0;
		opts->tsecr = tp->rx_opt.ts_recent;
		size += TCPOLEN_TSTAMP_ALIGNED;
	}

760 761
	eff_sacks = tp->rx_opt.num_sacks + tp->rx_opt.dsack;
	if (unlikely(eff_sacks)) {
A
Adam Langley 已提交
762 763
		const unsigned remaining = MAX_TCP_OPTION_SPACE - size;
		opts->num_sack_blocks =
764
			min_t(unsigned, eff_sacks,
A
Adam Langley 已提交
765 766 767 768 769 770 771
			      (remaining - TCPOLEN_SACK_BASE_ALIGNED) /
			      TCPOLEN_SACK_PERBLOCK);
		size += TCPOLEN_SACK_BASE_ALIGNED +
			opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK;
	}

	return size;
S
Stephen Hemminger 已提交
772
}
L
Linus Torvalds 已提交
773 774 775 776 777 778 779 780 781 782 783 784

/* This routine actually transmits TCP packets queued in by
 * tcp_do_sendmsg().  This is used by both the initial
 * transmission and possible later retransmissions.
 * All SKB's seen here are completely headerless.  It is our
 * job to build the TCP header, and pass the packet down to
 * IP so it can do the same plus pass the packet off to the
 * device.
 *
 * We are working here with either a clone of the original
 * SKB, or a fresh unique copy made by the retransmit engine.
 */
785 786
static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it,
			    gfp_t gfp_mask)
L
Linus Torvalds 已提交
787
{
788 789 790 791
	const struct inet_connection_sock *icsk = inet_csk(sk);
	struct inet_sock *inet;
	struct tcp_sock *tp;
	struct tcp_skb_cb *tcb;
A
Adam Langley 已提交
792 793
	struct tcp_out_options opts;
	unsigned tcp_options_size, tcp_header_size;
794
	struct tcp_md5sig_key *md5;
795 796 797 798 799 800 801 802
	struct tcphdr *th;
	int err;

	BUG_ON(!skb || !tcp_skb_pcount(skb));

	/* If congestion control is doing timestamping, we must
	 * take such a timestamp before we potentially clone/copy.
	 */
803
	if (icsk->icsk_ca_ops->flags & TCP_CONG_RTT_STAMP)
804 805 806 807 808 809 810 811 812 813
		__net_timestamp(skb);

	if (likely(clone_it)) {
		if (unlikely(skb_cloned(skb)))
			skb = pskb_copy(skb, gfp_mask);
		else
			skb = skb_clone(skb, gfp_mask);
		if (unlikely(!skb))
			return -ENOBUFS;
	}
L
Linus Torvalds 已提交
814

815 816 817
	inet = inet_sk(sk);
	tp = tcp_sk(sk);
	tcb = TCP_SKB_CB(skb);
A
Adam Langley 已提交
818
	memset(&opts, 0, sizeof(opts));
L
Linus Torvalds 已提交
819

A
Adam Langley 已提交
820 821 822 823 824 825
	if (unlikely(tcb->flags & TCPCB_FLAG_SYN))
		tcp_options_size = tcp_syn_options(sk, skb, &opts, &md5);
	else
		tcp_options_size = tcp_established_options(sk, skb, &opts,
							   &md5);
	tcp_header_size = tcp_options_size + sizeof(struct tcphdr);
826

827 828 829
	if (tcp_packets_in_flight(tp) == 0)
		tcp_ca_event(sk, CA_EVENT_TX_START);

830 831
	skb_push(skb, tcp_header_size);
	skb_reset_transport_header(skb);
832
	skb_set_owner_w(skb, sk);
833 834

	/* Build TCP header and checksum it. */
835
	th = tcp_hdr(skb);
E
Eric Dumazet 已提交
836 837
	th->source		= inet->inet_sport;
	th->dest		= inet->inet_dport;
838 839
	th->seq			= htonl(tcb->seq);
	th->ack_seq		= htonl(tp->rcv_nxt);
840
	*(((__be16 *)th) + 6)	= htons(((tcp_header_size >> 2) << 12) |
841 842 843 844 845 846
					tcb->flags);

	if (unlikely(tcb->flags & TCPCB_FLAG_SYN)) {
		/* RFC1323: The window in SYN & SYN/ACK segments
		 * is never scaled.
		 */
847
		th->window	= htons(min(tp->rcv_wnd, 65535U));
848 849 850 851 852
	} else {
		th->window	= htons(tcp_select_window(sk));
	}
	th->check		= 0;
	th->urg_ptr		= 0;
L
Linus Torvalds 已提交
853

I
Ilpo Järvinen 已提交
854
	/* The urg_mode check is necessary during a below snd_una win probe */
855 856 857 858 859 860 861 862
	if (unlikely(tcp_urg_mode(tp) && before(tcb->seq, tp->snd_up))) {
		if (before(tp->snd_up, tcb->seq + 0x10000)) {
			th->urg_ptr = htons(tp->snd_up - tcb->seq);
			th->urg = 1;
		} else if (after(tcb->seq + 0xFFFF, tp->snd_nxt)) {
			th->urg_ptr = 0xFFFF;
			th->urg = 1;
		}
863
	}
L
Linus Torvalds 已提交
864

865
	tcp_options_write((__be32 *)(th + 1), tp, &opts);
A
Adam Langley 已提交
866
	if (likely((tcb->flags & TCPCB_FLAG_SYN) == 0))
867
		TCP_ECN_send(sk, skb, tcp_header_size);
L
Linus Torvalds 已提交
868

869 870 871
#ifdef CONFIG_TCP_MD5SIG
	/* Calculate the MD5 hash, as we have all we need now */
	if (md5) {
A
Adam Langley 已提交
872
		sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
873
		tp->af_specific->calc_md5_hash(opts.hash_location,
874
					       md5, sk, NULL, skb);
875 876 877
	}
#endif

878
	icsk->icsk_af_ops->send_check(sk, skb->len, skb);
L
Linus Torvalds 已提交
879

880 881
	if (likely(tcb->flags & TCPCB_FLAG_ACK))
		tcp_event_ack_sent(sk, tcp_skb_pcount(skb));
L
Linus Torvalds 已提交
882

883 884
	if (skb->len != tcp_header_size)
		tcp_event_data_sent(tp, skb, sk);
L
Linus Torvalds 已提交
885

886
	if (after(tcb->end_seq, tp->snd_nxt) || tcb->seq == tcb->end_seq)
P
Pavel Emelyanov 已提交
887
		TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTSEGS);
L
Linus Torvalds 已提交
888

889
	err = icsk->icsk_af_ops->queue_xmit(skb, 0);
890
	if (likely(err <= 0))
891 892
		return err;

893
	tcp_enter_cwr(sk, 1);
894

895
	return net_xmit_eval(err);
L
Linus Torvalds 已提交
896 897
}

898
/* This routine just queues the buffer for sending.
L
Linus Torvalds 已提交
899 900 901 902 903 904 905 906 907 908 909
 *
 * NOTE: probe0 timer is not checked, do not forget tcp_push_pending_frames,
 * otherwise socket can stall.
 */
static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb)
{
	struct tcp_sock *tp = tcp_sk(sk);

	/* Advance write_seq and place onto the write_queue. */
	tp->write_seq = TCP_SKB_CB(skb)->end_seq;
	skb_header_release(skb);
910
	tcp_add_write_queue_tail(sk, skb);
911 912
	sk->sk_wmem_queued += skb->truesize;
	sk_mem_charge(sk, skb->truesize);
L
Linus Torvalds 已提交
913 914
}

915
/* Initialize TSO segments for a packet. */
916 917
static void tcp_set_skb_tso_segs(struct sock *sk, struct sk_buff *skb,
				 unsigned int mss_now)
918
{
919 920
	if (skb->len <= mss_now || !sk_can_gso(sk) ||
	    skb->ip_summed == CHECKSUM_NONE) {
921 922 923
		/* Avoid the costly divide in the normal
		 * non-TSO case.
		 */
924 925 926
		skb_shinfo(skb)->gso_segs = 1;
		skb_shinfo(skb)->gso_size = 0;
		skb_shinfo(skb)->gso_type = 0;
927
	} else {
I
Ilpo Järvinen 已提交
928
		skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss_now);
929
		skb_shinfo(skb)->gso_size = mss_now;
930
		skb_shinfo(skb)->gso_type = sk->sk_gso_type;
L
Linus Torvalds 已提交
931 932 933
	}
}

I
Ilpo Järvinen 已提交
934
/* When a modification to fackets out becomes necessary, we need to check
935
 * skb is counted to fackets_out or not.
I
Ilpo Järvinen 已提交
936
 */
937
static void tcp_adjust_fackets_out(struct sock *sk, struct sk_buff *skb,
I
Ilpo Järvinen 已提交
938 939
				   int decr)
{
940 941
	struct tcp_sock *tp = tcp_sk(sk);

942
	if (!tp->sacked_out || tcp_is_reno(tp))
I
Ilpo Järvinen 已提交
943 944
		return;

945
	if (after(tcp_highest_sack_seq(tp), TCP_SKB_CB(skb)->seq))
I
Ilpo Järvinen 已提交
946 947 948
		tp->fackets_out -= decr;
}

949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972
/* Pcount in the middle of the write queue got changed, we need to do various
 * tweaks to fix counters
 */
static void tcp_adjust_pcount(struct sock *sk, struct sk_buff *skb, int decr)
{
	struct tcp_sock *tp = tcp_sk(sk);

	tp->packets_out -= decr;

	if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)
		tp->sacked_out -= decr;
	if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS)
		tp->retrans_out -= decr;
	if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST)
		tp->lost_out -= decr;

	/* Reno case is special. Sigh... */
	if (tcp_is_reno(tp) && decr > 0)
		tp->sacked_out -= min_t(u32, tp->sacked_out, decr);

	tcp_adjust_fackets_out(sk, skb, decr);

	if (tp->lost_skb_hint &&
	    before(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(tp->lost_skb_hint)->seq) &&
973
	    (tcp_is_fack(tp) || (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)))
974 975 976 977 978
		tp->lost_cnt_hint -= decr;

	tcp_verify_left_out(tp);
}

L
Linus Torvalds 已提交
979 980
/* Function to create two new TCP segments.  Shrinks the given segment
 * to the specified size and appends a new segment with the rest of the
981
 * packet to the list.  This won't be called frequently, I hope.
L
Linus Torvalds 已提交
982 983
 * Remember, these are still headerless SKBs at this point.
 */
984 985
int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len,
		 unsigned int mss_now)
L
Linus Torvalds 已提交
986 987 988
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *buff;
989
	int nsize, old_factor;
990
	int nlen;
991
	u8 flags;
L
Linus Torvalds 已提交
992

993
	BUG_ON(len > skb->len);
994

L
Linus Torvalds 已提交
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
	nsize = skb_headlen(skb) - len;
	if (nsize < 0)
		nsize = 0;

	if (skb_cloned(skb) &&
	    skb_is_nonlinear(skb) &&
	    pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
		return -ENOMEM;

	/* Get a new skb... force flag on. */
	buff = sk_stream_alloc_skb(sk, nsize, GFP_ATOMIC);
	if (buff == NULL)
		return -ENOMEM; /* We'll just try again later. */
H
Herbert Xu 已提交
1008

1009 1010
	sk->sk_wmem_queued += buff->truesize;
	sk_mem_charge(sk, buff->truesize);
1011 1012 1013
	nlen = skb->len - len - nsize;
	buff->truesize += nlen;
	skb->truesize -= nlen;
L
Linus Torvalds 已提交
1014 1015 1016 1017 1018 1019 1020 1021

	/* Correct the sequence numbers. */
	TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len;
	TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq;
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq;

	/* PSH and FIN should only be set in the second packet. */
	flags = TCP_SKB_CB(skb)->flags;
1022
	TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN | TCPCB_FLAG_PSH);
L
Linus Torvalds 已提交
1023
	TCP_SKB_CB(buff)->flags = flags;
1024
	TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked;
L
Linus Torvalds 已提交
1025

1026
	if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_PARTIAL) {
L
Linus Torvalds 已提交
1027
		/* Copy and checksum data tail into the new buffer. */
1028 1029
		buff->csum = csum_partial_copy_nocheck(skb->data + len,
						       skb_put(buff, nsize),
L
Linus Torvalds 已提交
1030 1031 1032 1033 1034 1035
						       nsize, 0);

		skb_trim(skb, len);

		skb->csum = csum_block_sub(skb->csum, buff->csum, len);
	} else {
1036
		skb->ip_summed = CHECKSUM_PARTIAL;
L
Linus Torvalds 已提交
1037 1038 1039 1040 1041 1042 1043 1044 1045
		skb_split(skb, buff, len);
	}

	buff->ip_summed = skb->ip_summed;

	/* Looks stupid, but our code really uses when of
	 * skbs, which it never sent before. --ANK
	 */
	TCP_SKB_CB(buff)->when = TCP_SKB_CB(skb)->when;
1046
	buff->tstamp = skb->tstamp;
L
Linus Torvalds 已提交
1047

1048 1049
	old_factor = tcp_skb_pcount(skb);

L
Linus Torvalds 已提交
1050
	/* Fix up tso_factor for both original and new SKB.  */
1051 1052
	tcp_set_skb_tso_segs(sk, skb, mss_now);
	tcp_set_skb_tso_segs(sk, buff, mss_now);
L
Linus Torvalds 已提交
1053

1054 1055 1056
	/* If this packet has been sent out already, we must
	 * adjust the various packet counters.
	 */
1057
	if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq)) {
1058 1059
		int diff = old_factor - tcp_skb_pcount(skb) -
			tcp_skb_pcount(buff);
L
Linus Torvalds 已提交
1060

1061 1062
		if (diff)
			tcp_adjust_pcount(sk, skb, diff);
L
Linus Torvalds 已提交
1063 1064 1065
	}

	/* Link BUFF into the send queue. */
1066
	skb_header_release(buff);
1067
	tcp_insert_write_queue_after(skb, buff, sk);
L
Linus Torvalds 已提交
1068 1069 1070 1071 1072 1073 1074 1075

	return 0;
}

/* This is similar to __pskb_pull_head() (it will go to core/skbuff.c
 * eventually). The difference is that pulled data not copied, but
 * immediately discarded.
 */
1076
static void __pskb_trim_head(struct sk_buff *skb, int len)
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081
{
	int i, k, eat;

	eat = len;
	k = 0;
1082
	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
L
Linus Torvalds 已提交
1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
		if (skb_shinfo(skb)->frags[i].size <= eat) {
			put_page(skb_shinfo(skb)->frags[i].page);
			eat -= skb_shinfo(skb)->frags[i].size;
		} else {
			skb_shinfo(skb)->frags[k] = skb_shinfo(skb)->frags[i];
			if (eat) {
				skb_shinfo(skb)->frags[k].page_offset += eat;
				skb_shinfo(skb)->frags[k].size -= eat;
				eat = 0;
			}
			k++;
		}
	}
	skb_shinfo(skb)->nr_frags = k;

1098
	skb_reset_tail_pointer(skb);
L
Linus Torvalds 已提交
1099 1100 1101 1102
	skb->data_len -= len;
	skb->len = skb->data_len;
}

1103
/* Remove acked data from a packet in the transmit queue. */
L
Linus Torvalds 已提交
1104 1105
int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len)
{
1106
	if (skb_cloned(skb) && pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
L
Linus Torvalds 已提交
1107 1108
		return -ENOMEM;

1109 1110
	/* If len == headlen, we avoid __skb_pull to preserve alignment. */
	if (unlikely(len < skb_headlen(skb)))
L
Linus Torvalds 已提交
1111
		__skb_pull(skb, len);
1112 1113
	else
		__pskb_trim_head(skb, len - skb_headlen(skb));
L
Linus Torvalds 已提交
1114 1115

	TCP_SKB_CB(skb)->seq += len;
1116
	skb->ip_summed = CHECKSUM_PARTIAL;
L
Linus Torvalds 已提交
1117 1118 1119

	skb->truesize	     -= len;
	sk->sk_wmem_queued   -= len;
1120
	sk_mem_uncharge(sk, len);
L
Linus Torvalds 已提交
1121 1122 1123 1124 1125 1126
	sock_set_flag(sk, SOCK_QUEUE_SHRUNK);

	/* Any change of skb->len requires recalculation of tso
	 * factor and mss.
	 */
	if (tcp_skb_pcount(skb) > 1)
I
Ilpo Järvinen 已提交
1127
		tcp_set_skb_tso_segs(sk, skb, tcp_current_mss(sk));
L
Linus Torvalds 已提交
1128 1129 1130 1131

	return 0;
}

1132
/* Calculate MSS. Not accounting for SACKs here.  */
J
John Heffner 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175
int tcp_mtu_to_mss(struct sock *sk, int pmtu)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct inet_connection_sock *icsk = inet_csk(sk);
	int mss_now;

	/* Calculate base mss without TCP options:
	   It is MMS_S - sizeof(tcphdr) of rfc1122
	 */
	mss_now = pmtu - icsk->icsk_af_ops->net_header_len - sizeof(struct tcphdr);

	/* Clamp it (mss_clamp does not include tcp options) */
	if (mss_now > tp->rx_opt.mss_clamp)
		mss_now = tp->rx_opt.mss_clamp;

	/* Now subtract optional transport overhead */
	mss_now -= icsk->icsk_ext_hdr_len;

	/* Then reserve room for full set of TCP options and 8 bytes of data */
	if (mss_now < 48)
		mss_now = 48;

	/* Now subtract TCP options size, not including SACKs */
	mss_now -= tp->tcp_header_len - sizeof(struct tcphdr);

	return mss_now;
}

/* Inverse of above */
int tcp_mss_to_mtu(struct sock *sk, int mss)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct inet_connection_sock *icsk = inet_csk(sk);
	int mtu;

	mtu = mss +
	      tp->tcp_header_len +
	      icsk->icsk_ext_hdr_len +
	      icsk->icsk_af_ops->net_header_len;

	return mtu;
}

1176
/* MTU probing init per socket */
J
John Heffner 已提交
1177 1178 1179 1180 1181 1182 1183
void tcp_mtup_init(struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct inet_connection_sock *icsk = inet_csk(sk);

	icsk->icsk_mtup.enabled = sysctl_tcp_mtu_probing > 1;
	icsk->icsk_mtup.search_high = tp->rx_opt.mss_clamp + sizeof(struct tcphdr) +
1184
			       icsk->icsk_af_ops->net_header_len;
J
John Heffner 已提交
1185 1186 1187 1188
	icsk->icsk_mtup.search_low = tcp_mss_to_mtu(sk, sysctl_tcp_base_mss);
	icsk->icsk_mtup.probe_size = 0;
}

L
Linus Torvalds 已提交
1189 1190 1191 1192 1193 1194
/* This function synchronize snd mss to current pmtu/exthdr set.

   tp->rx_opt.user_mss is mss set by user by TCP_MAXSEG. It does NOT counts
   for TCP options, but includes only bare TCP header.

   tp->rx_opt.mss_clamp is mss negotiated at connection setup.
S
Stephen Hemminger 已提交
1195
   It is minimum of user_mss and mss received with SYN.
L
Linus Torvalds 已提交
1196 1197
   It also does not include TCP options.

1198
   inet_csk(sk)->icsk_pmtu_cookie is last pmtu, seen by this function.
L
Linus Torvalds 已提交
1199 1200 1201 1202 1203 1204 1205 1206 1207

   tp->mss_cache is current effective sending mss, including
   all tcp options except for SACKs. It is evaluated,
   taking into account current pmtu, but never exceeds
   tp->rx_opt.mss_clamp.

   NOTE1. rfc1122 clearly states that advertised MSS
   DOES NOT include either tcp or ip options.

1208 1209
   NOTE2. inet_csk(sk)->icsk_pmtu_cookie and tp->mss_cache
   are READ ONLY outside this function.		--ANK (980731)
L
Linus Torvalds 已提交
1210 1211 1212 1213
 */
unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
{
	struct tcp_sock *tp = tcp_sk(sk);
1214
	struct inet_connection_sock *icsk = inet_csk(sk);
J
John Heffner 已提交
1215
	int mss_now;
L
Linus Torvalds 已提交
1216

J
John Heffner 已提交
1217 1218
	if (icsk->icsk_mtup.search_high > pmtu)
		icsk->icsk_mtup.search_high = pmtu;
L
Linus Torvalds 已提交
1219

J
John Heffner 已提交
1220
	mss_now = tcp_mtu_to_mss(sk, pmtu);
1221
	mss_now = tcp_bound_to_half_wnd(tp, mss_now);
L
Linus Torvalds 已提交
1222 1223

	/* And store cached results */
1224
	icsk->icsk_pmtu_cookie = pmtu;
J
John Heffner 已提交
1225 1226
	if (icsk->icsk_mtup.enabled)
		mss_now = min(mss_now, tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_low));
1227
	tp->mss_cache = mss_now;
L
Linus Torvalds 已提交
1228 1229 1230 1231 1232 1233 1234

	return mss_now;
}

/* Compute the current effective MSS, taking SACKs and IP options,
 * and even PMTU discovery events into account.
 */
I
Ilpo Järvinen 已提交
1235
unsigned int tcp_current_mss(struct sock *sk)
L
Linus Torvalds 已提交
1236 1237 1238
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct dst_entry *dst = __sk_dst_get(sk);
1239
	u32 mss_now;
A
Adam Langley 已提交
1240 1241 1242
	unsigned header_len;
	struct tcp_out_options opts;
	struct tcp_md5sig_key *md5;
1243 1244 1245

	mss_now = tp->mss_cache;

L
Linus Torvalds 已提交
1246 1247
	if (dst) {
		u32 mtu = dst_mtu(dst);
1248
		if (mtu != inet_csk(sk)->icsk_pmtu_cookie)
L
Linus Torvalds 已提交
1249 1250 1251
			mss_now = tcp_sync_mss(sk, mtu);
	}

A
Adam Langley 已提交
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261
	header_len = tcp_established_options(sk, NULL, &opts, &md5) +
		     sizeof(struct tcphdr);
	/* The mss_cache is sized based on tp->tcp_header_len, which assumes
	 * some common options. If this is an odd packet (because we have SACK
	 * blocks etc) then our calculated header_len will be different, and
	 * we have to adjust mss_now correspondingly */
	if (header_len != tp->tcp_header_len) {
		int delta = (int) header_len - tp->tcp_header_len;
		mss_now -= delta;
	}
1262

L
Linus Torvalds 已提交
1263 1264 1265
	return mss_now;
}

1266
/* Congestion window validation. (RFC2861) */
1267
static void tcp_cwnd_validate(struct sock *sk)
1268
{
1269
	struct tcp_sock *tp = tcp_sk(sk);
1270

1271
	if (tp->packets_out >= tp->snd_cwnd) {
1272 1273 1274 1275 1276 1277 1278 1279
		/* Network is feed fully. */
		tp->snd_cwnd_used = 0;
		tp->snd_cwnd_stamp = tcp_time_stamp;
	} else {
		/* Network starves. */
		if (tp->packets_out > tp->snd_cwnd_used)
			tp->snd_cwnd_used = tp->packets_out;

1280 1281
		if (sysctl_tcp_slow_start_after_idle &&
		    (s32)(tcp_time_stamp - tp->snd_cwnd_stamp) >= inet_csk(sk)->icsk_rto)
1282 1283 1284 1285
			tcp_cwnd_application_limited(sk);
	}
}

1286 1287 1288 1289
/* Returns the portion of skb which can be sent right away without
 * introducing MSS oddities to segment boundaries. In rare cases where
 * mss_now != mss_cache, we will request caller to create a small skb
 * per input skb which could be mostly avoided here (if desired).
1290 1291 1292 1293 1294 1295 1296
 *
 * We explicitly want to create a request for splitting write queue tail
 * to a small skb for Nagle purposes while avoiding unnecessary modulos,
 * thus all the complexity (cwnd_len is always MSS multiple which we
 * return whenever allowed by the other factors). Basically we need the
 * modulo only when the receiver window alone is the limiting factor or
 * when we would be allowed to send the split-due-to-Nagle skb fully.
1297 1298
 */
static unsigned int tcp_mss_split_point(struct sock *sk, struct sk_buff *skb,
1299
					unsigned int mss_now, unsigned int cwnd)
1300
{
1301 1302
	struct tcp_sock *tp = tcp_sk(sk);
	u32 needed, window, cwnd_len;
1303

1304
	window = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
1305
	cwnd_len = mss_now * cwnd;
1306 1307 1308 1309

	if (likely(cwnd_len <= window && skb != tcp_write_queue_tail(sk)))
		return cwnd_len;

1310 1311
	needed = min(skb->len, window);

1312
	if (cwnd_len <= needed)
1313 1314 1315
		return cwnd_len;

	return needed - needed % mss_now;
1316 1317 1318 1319 1320
}

/* Can at least one segment of SKB be sent right now, according to the
 * congestion window rules?  If so, return how many segments are allowed.
 */
1321 1322
static inline unsigned int tcp_cwnd_test(struct tcp_sock *tp,
					 struct sk_buff *skb)
1323 1324 1325 1326
{
	u32 in_flight, cwnd;

	/* Don't be strict about the congestion window for the final FIN.  */
1327 1328
	if ((TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN) &&
	    tcp_skb_pcount(skb) == 1)
1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
		return 1;

	in_flight = tcp_packets_in_flight(tp);
	cwnd = tp->snd_cwnd;
	if (in_flight < cwnd)
		return (cwnd - in_flight);

	return 0;
}

1339 1340
/* Intialize TSO state of a skb.
 * This must be invoked the first time we consider transmitting
1341 1342
 * SKB onto the wire.
 */
1343 1344
static int tcp_init_tso_segs(struct sock *sk, struct sk_buff *skb,
			     unsigned int mss_now)
1345 1346 1347
{
	int tso_segs = tcp_skb_pcount(skb);

1348
	if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now)) {
1349
		tcp_set_skb_tso_segs(sk, skb, mss_now);
1350 1351 1352 1353 1354
		tso_segs = tcp_skb_pcount(skb);
	}
	return tso_segs;
}

1355
/* Minshall's variant of the Nagle send check. */
1356 1357
static inline int tcp_minshall_check(const struct tcp_sock *tp)
{
1358
	return after(tp->snd_sml, tp->snd_una) &&
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369
		!after(tp->snd_sml, tp->snd_nxt);
}

/* Return 0, if packet can be sent now without violation Nagle's rules:
 * 1. It is full sized.
 * 2. Or it contains FIN. (already checked by caller)
 * 3. Or TCP_NODELAY was set.
 * 4. Or TCP_CORK is not set, and all sent packets are ACKed.
 *    With Minshall's modification: all sent small packets are ACKed.
 */
static inline int tcp_nagle_check(const struct tcp_sock *tp,
1370
				  const struct sk_buff *skb,
1371 1372 1373
				  unsigned mss_now, int nonagle)
{
	return (skb->len < mss_now &&
1374 1375
		((nonagle & TCP_NAGLE_CORK) ||
		 (!nonagle && tp->packets_out && tcp_minshall_check(tp))));
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
}

/* Return non-zero if the Nagle test allows this packet to be
 * sent now.
 */
static inline int tcp_nagle_test(struct tcp_sock *tp, struct sk_buff *skb,
				 unsigned int cur_mss, int nonagle)
{
	/* Nagle rule does not apply to frames, which sit in the middle of the
	 * write_queue (they have no chances to get new data).
	 *
	 * This is implemented in the callers, where they modify the 'nonagle'
	 * argument based upon the location of SKB in the send queue.
	 */
	if (nonagle & TCP_NAGLE_PUSH)
		return 1;

1393 1394 1395
	/* Don't use the nagle rule for urgent data (or for the final FIN).
	 * Nagle can be ignored during F-RTO too (see RFC4138).
	 */
I
Ilpo Järvinen 已提交
1396
	if (tcp_urg_mode(tp) || (tp->frto_counter == 2) ||
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406
	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN))
		return 1;

	if (!tcp_nagle_check(tp, skb, cur_mss, nonagle))
		return 1;

	return 0;
}

/* Does at least the first segment of SKB fit into the send window? */
1407 1408
static inline int tcp_snd_wnd_test(struct tcp_sock *tp, struct sk_buff *skb,
				   unsigned int cur_mss)
1409 1410 1411 1412 1413 1414
{
	u32 end_seq = TCP_SKB_CB(skb)->end_seq;

	if (skb->len > cur_mss)
		end_seq = TCP_SKB_CB(skb)->seq + cur_mss;

1415
	return !after(end_seq, tcp_wnd_end(tp));
1416 1417
}

1418
/* This checks if the data bearing packet SKB (usually tcp_send_head(sk))
1419 1420 1421 1422 1423 1424 1425 1426 1427
 * should be put on the wire right now.  If so, it returns the number of
 * packets allowed by the congestion window.
 */
static unsigned int tcp_snd_test(struct sock *sk, struct sk_buff *skb,
				 unsigned int cur_mss, int nonagle)
{
	struct tcp_sock *tp = tcp_sk(sk);
	unsigned int cwnd_quota;

1428
	tcp_init_tso_segs(sk, skb, cur_mss);
1429 1430 1431 1432 1433

	if (!tcp_nagle_test(tp, skb, cur_mss, nonagle))
		return 0;

	cwnd_quota = tcp_cwnd_test(tp, skb);
1434
	if (cwnd_quota && !tcp_snd_wnd_test(tp, skb, cur_mss))
1435 1436 1437 1438 1439
		cwnd_quota = 0;

	return cwnd_quota;
}

1440
/* Test if sending is allowed right now. */
1441
int tcp_may_send_now(struct sock *sk)
1442
{
1443
	struct tcp_sock *tp = tcp_sk(sk);
1444
	struct sk_buff *skb = tcp_send_head(sk);
1445 1446

	return (skb &&
I
Ilpo Järvinen 已提交
1447
		tcp_snd_test(sk, skb, tcp_current_mss(sk),
1448
			     (tcp_skb_is_last(sk, skb) ?
1449
			      tp->nonagle : TCP_NAGLE_PUSH)));
1450 1451 1452 1453 1454 1455 1456 1457 1458
}

/* Trim TSO SKB to LEN bytes, put the remaining data into a new packet
 * which is put after SKB on the list.  It is very much like
 * tcp_fragment() except that it may make several kinds of assumptions
 * in order to speed up the splitting operation.  In particular, we
 * know that all the data is in scatter-gather pages, and that the
 * packet has never been sent out before (and thus is not cloned).
 */
1459 1460
static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len,
			unsigned int mss_now)
1461 1462 1463
{
	struct sk_buff *buff;
	int nlen = skb->len - len;
1464
	u8 flags;
1465 1466

	/* All of a TSO frame must be composed of paged data.  */
1467 1468
	if (skb->len != skb->data_len)
		return tcp_fragment(sk, skb, len, mss_now);
1469

1470
	buff = sk_stream_alloc_skb(sk, 0, GFP_ATOMIC);
1471 1472 1473
	if (unlikely(buff == NULL))
		return -ENOMEM;

1474 1475
	sk->sk_wmem_queued += buff->truesize;
	sk_mem_charge(sk, buff->truesize);
1476
	buff->truesize += nlen;
1477 1478 1479 1480 1481 1482 1483 1484 1485
	skb->truesize -= nlen;

	/* Correct the sequence numbers. */
	TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len;
	TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq;
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq;

	/* PSH and FIN should only be set in the second packet. */
	flags = TCP_SKB_CB(skb)->flags;
1486
	TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN | TCPCB_FLAG_PSH);
1487 1488 1489 1490 1491
	TCP_SKB_CB(buff)->flags = flags;

	/* This packet was never sent out yet, so no SACK bits. */
	TCP_SKB_CB(buff)->sacked = 0;

1492
	buff->ip_summed = skb->ip_summed = CHECKSUM_PARTIAL;
1493 1494 1495
	skb_split(skb, buff, len);

	/* Fix up tso_factor for both original and new SKB.  */
1496 1497
	tcp_set_skb_tso_segs(sk, skb, mss_now);
	tcp_set_skb_tso_segs(sk, buff, mss_now);
1498 1499 1500

	/* Link BUFF into the send queue. */
	skb_header_release(buff);
1501
	tcp_insert_write_queue_after(skb, buff, sk);
1502 1503 1504 1505 1506 1507 1508 1509 1510

	return 0;
}

/* Try to defer sending, if possible, in order to minimize the amount
 * of TSO splitting we do.  View it as a kind of TSO Nagle test.
 *
 * This algorithm is from John Heffner.
 */
1511
static int tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb)
1512
{
1513
	struct tcp_sock *tp = tcp_sk(sk);
1514
	const struct inet_connection_sock *icsk = inet_csk(sk);
1515 1516 1517
	u32 send_win, cong_win, limit, in_flight;

	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
J
John Heffner 已提交
1518
		goto send_now;
1519

1520
	if (icsk->icsk_ca_state != TCP_CA_Open)
J
John Heffner 已提交
1521 1522 1523
		goto send_now;

	/* Defer for less than two clock ticks. */
I
Ilpo Järvinen 已提交
1524
	if (tp->tso_deferred &&
1525
	    (((u32)jiffies << 1) >> 1) - (tp->tso_deferred >> 1) > 1)
J
John Heffner 已提交
1526
		goto send_now;
1527

1528 1529
	in_flight = tcp_packets_in_flight(tp);

1530
	BUG_ON(tcp_skb_pcount(skb) <= 1 || (tp->snd_cwnd <= in_flight));
1531

1532
	send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
1533 1534 1535 1536 1537 1538

	/* From in_flight test above, we know that cwnd > in_flight.  */
	cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache;

	limit = min(send_win, cong_win);

1539
	/* If a full-sized TSO skb can be sent, do it. */
1540
	if (limit >= sk->sk_gso_max_size)
J
John Heffner 已提交
1541
		goto send_now;
1542

1543 1544 1545 1546
	/* Middle in queue won't get any more data, full sendable already? */
	if ((skb != tcp_write_queue_tail(sk)) && (limit >= skb->len))
		goto send_now;

1547 1548 1549 1550 1551 1552 1553 1554
	if (sysctl_tcp_tso_win_divisor) {
		u32 chunk = min(tp->snd_wnd, tp->snd_cwnd * tp->mss_cache);

		/* If at least some fraction of a window is available,
		 * just use it.
		 */
		chunk /= sysctl_tcp_tso_win_divisor;
		if (limit >= chunk)
J
John Heffner 已提交
1555
			goto send_now;
1556 1557 1558 1559 1560 1561 1562
	} else {
		/* Different approach, try not to defer past a single
		 * ACK.  Receiver should ACK every other full sized
		 * frame, so if we have space for more than 3 frames
		 * then send now.
		 */
		if (limit > tcp_max_burst(tp) * tp->mss_cache)
J
John Heffner 已提交
1563
			goto send_now;
1564 1565 1566
	}

	/* Ok, it looks like it is advisable to defer.  */
1567
	tp->tso_deferred = 1 | (jiffies << 1);
J
John Heffner 已提交
1568

1569
	return 1;
J
John Heffner 已提交
1570 1571 1572 1573

send_now:
	tp->tso_deferred = 0;
	return 0;
1574 1575
}

J
John Heffner 已提交
1576
/* Create a new MTU probe if we are ready.
1577 1578 1579 1580
 * MTU probe is regularly attempting to increase the path MTU by
 * deliberately sending larger packets.  This discovers routing
 * changes resulting in larger path MTUs.
 *
J
John Heffner 已提交
1581 1582
 * Returns 0 if we should wait to probe (no cwnd available),
 *         1 if a probe was sent,
1583 1584
 *         -1 otherwise
 */
J
John Heffner 已提交
1585 1586 1587 1588 1589 1590 1591
static int tcp_mtu_probe(struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct inet_connection_sock *icsk = inet_csk(sk);
	struct sk_buff *skb, *nskb, *next;
	int len;
	int probe_size;
1592
	int size_needed;
J
John Heffner 已提交
1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603
	int copy;
	int mss_now;

	/* Not currently probing/verifying,
	 * not in recovery,
	 * have enough cwnd, and
	 * not SACKing (the variable headers throw things off) */
	if (!icsk->icsk_mtup.enabled ||
	    icsk->icsk_mtup.probe_size ||
	    inet_csk(sk)->icsk_ca_state != TCP_CA_Open ||
	    tp->snd_cwnd < 11 ||
1604
	    tp->rx_opt.num_sacks || tp->rx_opt.dsack)
J
John Heffner 已提交
1605 1606 1607
		return -1;

	/* Very simple search strategy: just double the MSS. */
I
Ilpo Järvinen 已提交
1608
	mss_now = tcp_current_mss(sk);
1609
	probe_size = 2 * tp->mss_cache;
1610
	size_needed = probe_size + (tp->reordering + 1) * tp->mss_cache;
J
John Heffner 已提交
1611 1612 1613 1614 1615 1616
	if (probe_size > tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_high)) {
		/* TODO: set timer for probe_converge_event */
		return -1;
	}

	/* Have enough data in the send queue to probe? */
1617
	if (tp->write_seq - tp->snd_nxt < size_needed)
J
John Heffner 已提交
1618 1619
		return -1;

1620 1621
	if (tp->snd_wnd < size_needed)
		return -1;
1622
	if (after(tp->snd_nxt + size_needed, tcp_wnd_end(tp)))
1623
		return 0;
J
John Heffner 已提交
1624

1625 1626 1627
	/* Do we need to wait to drain cwnd? With none in flight, don't stall */
	if (tcp_packets_in_flight(tp) + 2 > tp->snd_cwnd) {
		if (!tcp_packets_in_flight(tp))
J
John Heffner 已提交
1628 1629 1630 1631 1632 1633 1634 1635
			return -1;
		else
			return 0;
	}

	/* We're allowed to probe.  Build it now. */
	if ((nskb = sk_stream_alloc_skb(sk, probe_size, GFP_ATOMIC)) == NULL)
		return -1;
1636 1637
	sk->sk_wmem_queued += nskb->truesize;
	sk_mem_charge(sk, nskb->truesize);
J
John Heffner 已提交
1638

1639
	skb = tcp_send_head(sk);
J
John Heffner 已提交
1640 1641 1642 1643 1644 1645

	TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(skb)->seq;
	TCP_SKB_CB(nskb)->end_seq = TCP_SKB_CB(skb)->seq + probe_size;
	TCP_SKB_CB(nskb)->flags = TCPCB_FLAG_ACK;
	TCP_SKB_CB(nskb)->sacked = 0;
	nskb->csum = 0;
1646
	nskb->ip_summed = skb->ip_summed;
J
John Heffner 已提交
1647

1648 1649
	tcp_insert_write_queue_before(nskb, skb, sk);

J
John Heffner 已提交
1650
	len = 0;
1651
	tcp_for_write_queue_from_safe(skb, next, sk) {
J
John Heffner 已提交
1652 1653 1654 1655 1656
		copy = min_t(int, skb->len, probe_size - len);
		if (nskb->ip_summed)
			skb_copy_bits(skb, 0, skb_put(nskb, copy), copy);
		else
			nskb->csum = skb_copy_and_csum_bits(skb, 0,
1657 1658
							    skb_put(nskb, copy),
							    copy, nskb->csum);
J
John Heffner 已提交
1659 1660 1661 1662 1663

		if (skb->len <= copy) {
			/* We've eaten all the data from this skb.
			 * Throw it away. */
			TCP_SKB_CB(nskb)->flags |= TCP_SKB_CB(skb)->flags;
1664
			tcp_unlink_write_queue(skb, sk);
1665
			sk_wmem_free_skb(sk, skb);
J
John Heffner 已提交
1666 1667
		} else {
			TCP_SKB_CB(nskb)->flags |= TCP_SKB_CB(skb)->flags &
1668
						   ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
J
John Heffner 已提交
1669 1670
			if (!skb_shinfo(skb)->nr_frags) {
				skb_pull(skb, copy);
1671
				if (skb->ip_summed != CHECKSUM_PARTIAL)
1672 1673
					skb->csum = csum_partial(skb->data,
								 skb->len, 0);
J
John Heffner 已提交
1674 1675 1676 1677 1678 1679 1680 1681
			} else {
				__pskb_trim_head(skb, copy);
				tcp_set_skb_tso_segs(sk, skb, mss_now);
			}
			TCP_SKB_CB(skb)->seq += copy;
		}

		len += copy;
1682 1683 1684

		if (len >= probe_size)
			break;
J
John Heffner 已提交
1685 1686 1687 1688 1689 1690 1691 1692
	}
	tcp_init_tso_segs(sk, nskb, nskb->len);

	/* We're ready to send.  If this fails, the probe will
	 * be resegmented into mss-sized pieces by tcp_write_xmit(). */
	TCP_SKB_CB(nskb)->when = tcp_time_stamp;
	if (!tcp_transmit_skb(sk, nskb, 1, GFP_ATOMIC)) {
		/* Decrement cwnd here because we are sending
1693
		 * effectively two packets. */
J
John Heffner 已提交
1694
		tp->snd_cwnd--;
1695
		tcp_event_new_data_sent(sk, nskb);
J
John Heffner 已提交
1696 1697

		icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len);
1698 1699
		tp->mtu_probe.probe_seq_start = TCP_SKB_CB(nskb)->seq;
		tp->mtu_probe.probe_seq_end = TCP_SKB_CB(nskb)->end_seq;
J
John Heffner 已提交
1700 1701 1702 1703 1704 1705 1706

		return 1;
	}

	return -1;
}

L
Linus Torvalds 已提交
1707 1708 1709 1710
/* This routine writes packets to the network.  It advances the
 * send_head.  This happens as incoming acks open up the remote
 * window for us.
 *
1711 1712 1713 1714
 * LARGESEND note: !tcp_urg_mode is overkill, only frames between
 * snd_up-64k-mss .. snd_up cannot be large. However, taking into
 * account rare use of URG, this is not a big flaw.
 *
L
Linus Torvalds 已提交
1715 1716 1717
 * Returns 1, if no segments are in flight and we have queued segments, but
 * cannot send anything now because of SWS or another problem.
 */
1718 1719
static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
			  int push_one, gfp_t gfp)
L
Linus Torvalds 已提交
1720 1721
{
	struct tcp_sock *tp = tcp_sk(sk);
1722
	struct sk_buff *skb;
1723 1724
	unsigned int tso_segs, sent_pkts;
	int cwnd_quota;
J
John Heffner 已提交
1725
	int result;
L
Linus Torvalds 已提交
1726

1727
	sent_pkts = 0;
J
John Heffner 已提交
1728

1729 1730 1731 1732 1733 1734 1735 1736
	if (!push_one) {
		/* Do MTU probing. */
		result = tcp_mtu_probe(sk);
		if (!result) {
			return 0;
		} else if (result > 0) {
			sent_pkts = 1;
		}
J
John Heffner 已提交
1737 1738
	}

1739
	while ((skb = tcp_send_head(sk))) {
1740 1741
		unsigned int limit;

1742
		tso_segs = tcp_init_tso_segs(sk, skb, mss_now);
1743
		BUG_ON(!tso_segs);
1744

1745 1746 1747 1748 1749 1750 1751
		cwnd_quota = tcp_cwnd_test(tp, skb);
		if (!cwnd_quota)
			break;

		if (unlikely(!tcp_snd_wnd_test(tp, skb, mss_now)))
			break;

1752 1753 1754 1755 1756 1757
		if (tso_segs == 1) {
			if (unlikely(!tcp_nagle_test(tp, skb, mss_now,
						     (tcp_skb_is_last(sk, skb) ?
						      nonagle : TCP_NAGLE_PUSH))))
				break;
		} else {
1758
			if (!push_one && tcp_tso_should_defer(sk, skb))
1759 1760
				break;
		}
1761

1762
		limit = mss_now;
1763
		if (tso_segs > 1 && !tcp_urg_mode(tp))
1764 1765
			limit = tcp_mss_split_point(sk, skb, mss_now,
						    cwnd_quota);
L
Linus Torvalds 已提交
1766

1767 1768 1769 1770
		if (skb->len > limit &&
		    unlikely(tso_fragment(sk, skb, limit, mss_now)))
			break;

1771
		TCP_SKB_CB(skb)->when = tcp_time_stamp;
1772

1773
		if (unlikely(tcp_transmit_skb(sk, skb, 1, gfp)))
1774
			break;
L
Linus Torvalds 已提交
1775

1776 1777 1778
		/* Advance the send_head.  This one is sent out.
		 * This call will increment packets_out.
		 */
1779
		tcp_event_new_data_sent(sk, skb);
L
Linus Torvalds 已提交
1780

1781
		tcp_minshall_update(tp, mss_now, skb);
1782
		sent_pkts++;
1783 1784 1785

		if (push_one)
			break;
1786
	}
L
Linus Torvalds 已提交
1787

1788
	if (likely(sent_pkts)) {
1789
		tcp_cwnd_validate(sk);
1790
		return 0;
L
Linus Torvalds 已提交
1791
	}
1792
	return !tp->packets_out && tcp_send_head(sk);
L
Linus Torvalds 已提交
1793 1794
}

1795 1796 1797 1798
/* Push out any pending frames which were held back due to
 * TCP_CORK or attempt at coalescing tiny packets.
 * The socket must be locked by the caller.
 */
1799 1800
void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss,
			       int nonagle)
1801
{
1802
	struct sk_buff *skb = tcp_send_head(sk);
1803

1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
	if (!skb)
		return;

	/* If we are closed, the bytes will have to remain here.
	 * In time closedown will finish, we empty the write queue and
	 * all will be happy.
	 */
	if (unlikely(sk->sk_state == TCP_CLOSE))
		return;

1814
	if (tcp_write_xmit(sk, cur_mss, nonagle, 0, GFP_ATOMIC))
1815
		tcp_check_probe_timer(sk);
1816 1817
}

1818 1819 1820 1821 1822
/* Send _single_ skb sitting at the send head. This function requires
 * true push pending frames to setup probe timer etc.
 */
void tcp_push_one(struct sock *sk, unsigned int mss_now)
{
1823
	struct sk_buff *skb = tcp_send_head(sk);
1824 1825 1826

	BUG_ON(!skb || skb->len < mss_now);

1827
	tcp_write_xmit(sk, mss_now, TCP_NAGLE_PUSH, 1, sk->sk_allocation);
1828 1829
}

L
Linus Torvalds 已提交
1830 1831
/* This function returns the amount that we can raise the
 * usable window based on the following constraints
1832
 *
L
Linus Torvalds 已提交
1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850
 * 1. The window can never be shrunk once it is offered (RFC 793)
 * 2. We limit memory per socket
 *
 * RFC 1122:
 * "the suggested [SWS] avoidance algorithm for the receiver is to keep
 *  RECV.NEXT + RCV.WIN fixed until:
 *  RCV.BUFF - RCV.USER - RCV.WINDOW >= min(1/2 RCV.BUFF, MSS)"
 *
 * i.e. don't raise the right edge of the window until you can raise
 * it at least MSS bytes.
 *
 * Unfortunately, the recommended algorithm breaks header prediction,
 * since header prediction assumes th->window stays fixed.
 *
 * Strictly speaking, keeping th->window fixed violates the receiver
 * side SWS prevention criteria. The problem is that under this rule
 * a stream of single byte packets will cause the right side of the
 * window to always advance by a single byte.
1851
 *
L
Linus Torvalds 已提交
1852 1853
 * Of course, if the sender implements sender side SWS prevention
 * then this will not be a problem.
1854
 *
L
Linus Torvalds 已提交
1855
 * BSD seems to make the following compromise:
1856
 *
L
Linus Torvalds 已提交
1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883
 *	If the free space is less than the 1/4 of the maximum
 *	space available and the free space is less than 1/2 mss,
 *	then set the window to 0.
 *	[ Actually, bsd uses MSS and 1/4 of maximal _window_ ]
 *	Otherwise, just prevent the window from shrinking
 *	and from being larger than the largest representable value.
 *
 * This prevents incremental opening of the window in the regime
 * where TCP is limited by the speed of the reader side taking
 * data out of the TCP receive queue. It does nothing about
 * those cases where the window is constrained on the sender side
 * because the pipeline is full.
 *
 * BSD also seems to "accidentally" limit itself to windows that are a
 * multiple of MSS, at least until the free space gets quite small.
 * This would appear to be a side effect of the mbuf implementation.
 * Combining these two algorithms results in the observed behavior
 * of having a fixed window size at almost all times.
 *
 * Below we obtain similar behavior by forcing the offered window to
 * a multiple of the mss when it is feasible to do so.
 *
 * Note, we don't "adjust" for TIMESTAMP or SACK option bytes.
 * Regular options like TIMESTAMP are taken into account.
 */
u32 __tcp_select_window(struct sock *sk)
{
1884
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1885
	struct tcp_sock *tp = tcp_sk(sk);
S
Stephen Hemminger 已提交
1886
	/* MSS for the peer's data.  Previous versions used mss_clamp
L
Linus Torvalds 已提交
1887 1888 1889 1890 1891
	 * here.  I don't know if the value based on our guesses
	 * of peer's MSS is better for the performance.  It's more correct
	 * but may be worse for the performance because of rcv_mss
	 * fluctuations.  --SAW  1998/11/1
	 */
1892
	int mss = icsk->icsk_ack.rcv_mss;
L
Linus Torvalds 已提交
1893 1894 1895 1896 1897
	int free_space = tcp_space(sk);
	int full_space = min_t(int, tp->window_clamp, tcp_full_space(sk));
	int window;

	if (mss > full_space)
1898
		mss = full_space;
L
Linus Torvalds 已提交
1899

1900
	if (free_space < (full_space >> 1)) {
1901
		icsk->icsk_ack.quick = 0;
L
Linus Torvalds 已提交
1902 1903

		if (tcp_memory_pressure)
1904 1905
			tp->rcv_ssthresh = min(tp->rcv_ssthresh,
					       4U * tp->advmss);
L
Linus Torvalds 已提交
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937

		if (free_space < mss)
			return 0;
	}

	if (free_space > tp->rcv_ssthresh)
		free_space = tp->rcv_ssthresh;

	/* Don't do rounding if we are using window scaling, since the
	 * scaled window will not line up with the MSS boundary anyway.
	 */
	window = tp->rcv_wnd;
	if (tp->rx_opt.rcv_wscale) {
		window = free_space;

		/* Advertise enough space so that it won't get scaled away.
		 * Import case: prevent zero window announcement if
		 * 1<<rcv_wscale > mss.
		 */
		if (((window >> tp->rx_opt.rcv_wscale) << tp->rx_opt.rcv_wscale) != window)
			window = (((window >> tp->rx_opt.rcv_wscale) + 1)
				  << tp->rx_opt.rcv_wscale);
	} else {
		/* Get the largest window that is a nice multiple of mss.
		 * Window clamp already applied above.
		 * If our current window offering is within 1 mss of the
		 * free space we just keep it. This prevents the divide
		 * and multiply from happening most of the time.
		 * We also don't do any window rounding when the free space
		 * is too small.
		 */
		if (window <= free_space - mss || window > free_space)
1938
			window = (free_space / mss) * mss;
1939
		else if (mss == full_space &&
1940
			 free_space > window + (full_space >> 1))
1941
			window = free_space;
L
Linus Torvalds 已提交
1942 1943 1944 1945 1946
	}

	return window;
}

1947 1948
/* Collapses two adjacent SKB's during retransmission. */
static void tcp_collapse_retrans(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
1949 1950
{
	struct tcp_sock *tp = tcp_sk(sk);
1951
	struct sk_buff *next_skb = tcp_write_queue_next(sk, skb);
1952
	int skb_size, next_skb_size;
L
Linus Torvalds 已提交
1953

1954 1955
	skb_size = skb->len;
	next_skb_size = next_skb->len;
L
Linus Torvalds 已提交
1956

1957
	BUG_ON(tcp_skb_pcount(skb) != 1 || tcp_skb_pcount(next_skb) != 1);
1958

1959
	tcp_highest_sack_combine(sk, next_skb, skb);
L
Linus Torvalds 已提交
1960

1961
	tcp_unlink_write_queue(next_skb, sk);
L
Linus Torvalds 已提交
1962

1963 1964
	skb_copy_from_linear_data(next_skb, skb_put(skb, next_skb_size),
				  next_skb_size);
L
Linus Torvalds 已提交
1965

1966 1967
	if (next_skb->ip_summed == CHECKSUM_PARTIAL)
		skb->ip_summed = CHECKSUM_PARTIAL;
L
Linus Torvalds 已提交
1968

1969 1970
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		skb->csum = csum_block_add(skb->csum, next_skb->csum, skb_size);
L
Linus Torvalds 已提交
1971

1972 1973
	/* Update sequence range on original skb. */
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;
L
Linus Torvalds 已提交
1974

1975 1976
	/* Merge over control information. This moves PSH/FIN etc. over */
	TCP_SKB_CB(skb)->flags |= TCP_SKB_CB(next_skb)->flags;
1977 1978 1979 1980 1981 1982 1983

	/* All done, get rid of second SKB and account for it so
	 * packet counting does not break.
	 */
	TCP_SKB_CB(skb)->sacked |= TCP_SKB_CB(next_skb)->sacked & TCPCB_EVER_RETRANS;

	/* changed transmit queue under us so clear hints */
1984 1985 1986
	tcp_clear_retrans_hints_partial(tp);
	if (next_skb == tp->retransmit_skb_hint)
		tp->retransmit_skb_hint = skb;
1987

1988 1989
	tcp_adjust_pcount(sk, next_skb, tcp_skb_pcount(next_skb));

1990
	sk_wmem_free_skb(sk, next_skb);
L
Linus Torvalds 已提交
1991 1992
}

1993
/* Check if coalescing SKBs is legal. */
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
static int tcp_can_collapse(struct sock *sk, struct sk_buff *skb)
{
	if (tcp_skb_pcount(skb) > 1)
		return 0;
	/* TODO: SACK collapsing could be used to remove this condition */
	if (skb_shinfo(skb)->nr_frags != 0)
		return 0;
	if (skb_cloned(skb))
		return 0;
	if (skb == tcp_send_head(sk))
		return 0;
	/* Some heurestics for collapsing over SACK'd could be invented */
	if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)
		return 0;

	return 1;
}

2012 2013 2014
/* Collapse packets in the retransmit queue to make to create
 * less packets on the wire. This is only done on retransmission.
 */
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052
static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *to,
				     int space)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb = to, *tmp;
	int first = 1;

	if (!sysctl_tcp_retrans_collapse)
		return;
	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_SYN)
		return;

	tcp_for_write_queue_from_safe(skb, tmp, sk) {
		if (!tcp_can_collapse(sk, skb))
			break;

		space -= skb->len;

		if (first) {
			first = 0;
			continue;
		}

		if (space < 0)
			break;
		/* Punt if not enough space exists in the first SKB for
		 * the data in the second
		 */
		if (skb->len > skb_tailroom(to))
			break;

		if (after(TCP_SKB_CB(skb)->end_seq, tcp_wnd_end(tp)))
			break;

		tcp_collapse_retrans(sk, to);
	}
}

L
Linus Torvalds 已提交
2053 2054 2055 2056 2057 2058 2059
/* This retransmits one SKB.  Policy decisions and retransmit queue
 * state updates are done by the caller.  Returns non-zero if an
 * error occurred which prevented the send.
 */
int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
{
	struct tcp_sock *tp = tcp_sk(sk);
J
John Heffner 已提交
2060
	struct inet_connection_sock *icsk = inet_csk(sk);
2061
	unsigned int cur_mss;
L
Linus Torvalds 已提交
2062 2063
	int err;

J
John Heffner 已提交
2064 2065 2066 2067 2068
	/* Inconslusive MTU probe */
	if (icsk->icsk_mtup.probe_size) {
		icsk->icsk_mtup.probe_size = 0;
	}

L
Linus Torvalds 已提交
2069
	/* Do not sent more than we queued. 1/4 is reserved for possible
S
Stephen Hemminger 已提交
2070
	 * copying overhead: fragmentation, tunneling, mangling etc.
L
Linus Torvalds 已提交
2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082
	 */
	if (atomic_read(&sk->sk_wmem_alloc) >
	    min(sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), sk->sk_sndbuf))
		return -EAGAIN;

	if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) {
		if (before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))
			BUG();
		if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq))
			return -ENOMEM;
	}

2083 2084 2085
	if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk))
		return -EHOSTUNREACH; /* Routing failure or similar. */

I
Ilpo Järvinen 已提交
2086
	cur_mss = tcp_current_mss(sk);
2087

L
Linus Torvalds 已提交
2088 2089 2090 2091 2092
	/* If receiver has shrunk his window, and skb is out of
	 * new window, do not retransmit it. The exception is the
	 * case, when window is shrunk to zero. In this case
	 * our retransmit serves as a zero window probe.
	 */
2093 2094
	if (!before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp)) &&
	    TCP_SKB_CB(skb)->seq != tp->snd_una)
L
Linus Torvalds 已提交
2095 2096 2097
		return -EAGAIN;

	if (skb->len > cur_mss) {
2098
		if (tcp_fragment(sk, skb, cur_mss, cur_mss))
L
Linus Torvalds 已提交
2099
			return -ENOMEM; /* We'll try again later. */
2100
	} else {
2101 2102 2103 2104 2105 2106
		int oldpcount = tcp_skb_pcount(skb);

		if (unlikely(oldpcount > 1)) {
			tcp_init_tso_segs(sk, skb, cur_mss);
			tcp_adjust_pcount(sk, skb, oldpcount - tcp_skb_pcount(skb));
		}
L
Linus Torvalds 已提交
2107 2108
	}

2109
	tcp_retrans_try_collapse(sk, skb, cur_mss);
L
Linus Torvalds 已提交
2110 2111 2112 2113 2114

	/* Some Solaris stacks overoptimize and ignore the FIN on a
	 * retransmit when old data is attached.  So strip it off
	 * since it is cheap to do so and saves bytes on the network.
	 */
S
Stephen Hemminger 已提交
2115 2116 2117
	if (skb->len > 0 &&
	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN) &&
	    tp->snd_una == (TCP_SKB_CB(skb)->end_seq - 1)) {
L
Linus Torvalds 已提交
2118
		if (!pskb_trim(skb, 0)) {
2119 2120 2121
			/* Reuse, even though it does some unnecessary work */
			tcp_init_nondata_skb(skb, TCP_SKB_CB(skb)->end_seq - 1,
					     TCP_SKB_CB(skb)->flags);
L
Linus Torvalds 已提交
2122 2123 2124 2125 2126 2127 2128 2129 2130
			skb->ip_summed = CHECKSUM_NONE;
		}
	}

	/* Make a copy, if the first transmission SKB clone we made
	 * is still in somebody's hands, else make a clone.
	 */
	TCP_SKB_CB(skb)->when = tcp_time_stamp;

2131
	err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
L
Linus Torvalds 已提交
2132 2133 2134

	if (err == 0) {
		/* Update global TCP statistics. */
P
Pavel Emelyanov 已提交
2135
		TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS);
L
Linus Torvalds 已提交
2136 2137 2138 2139

		tp->total_retrans++;

#if FASTRETRANS_DEBUG > 0
2140
		if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS) {
L
Linus Torvalds 已提交
2141 2142 2143 2144
			if (net_ratelimit())
				printk(KERN_DEBUG "retrans_out leaked.\n");
		}
#endif
2145 2146
		if (!tp->retrans_out)
			tp->lost_retrans_low = tp->snd_nxt;
L
Linus Torvalds 已提交
2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163
		TCP_SKB_CB(skb)->sacked |= TCPCB_RETRANS;
		tp->retrans_out += tcp_skb_pcount(skb);

		/* Save stamp of the first retransmit. */
		if (!tp->retrans_stamp)
			tp->retrans_stamp = TCP_SKB_CB(skb)->when;

		tp->undo_retrans++;

		/* snd_nxt is stored to detect loss of retransmitted segment,
		 * see tcp_input.c tcp_sacktag_write_queue().
		 */
		TCP_SKB_CB(skb)->ack_seq = tp->snd_nxt;
	}
	return err;
}

2164 2165 2166
/* Check if we forward retransmits are possible in the current
 * window/congestion state.
 */
2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193
static int tcp_can_forward_retransmit(struct sock *sk)
{
	const struct inet_connection_sock *icsk = inet_csk(sk);
	struct tcp_sock *tp = tcp_sk(sk);

	/* Forward retransmissions are possible only during Recovery. */
	if (icsk->icsk_ca_state != TCP_CA_Recovery)
		return 0;

	/* No forward retransmissions in Reno are possible. */
	if (tcp_is_reno(tp))
		return 0;

	/* Yeah, we have to make difficult choice between forward transmission
	 * and retransmission... Both ways have their merits...
	 *
	 * For now we do not retransmit anything, while we have some new
	 * segments to send. In the other cases, follow rule 3 for
	 * NextSeg() specified in RFC3517.
	 */

	if (tcp_may_send_now(sk))
		return 0;

	return 1;
}

L
Linus Torvalds 已提交
2194 2195 2196 2197 2198 2199 2200 2201 2202 2203
/* This gets called after a retransmit timeout, and the initially
 * retransmitted data is acknowledged.  It tries to continue
 * resending the rest of the retransmit queue, until either
 * we've sent it all or the congestion window limit is reached.
 * If doing SACK, the first ACK which comes back for a timeout
 * based retransmit packet might feed us FACK information again.
 * If so, we use it to avoid unnecessarily retransmissions.
 */
void tcp_xmit_retransmit_queue(struct sock *sk)
{
2204
	const struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
2205 2206
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;
2207
	struct sk_buff *hole = NULL;
2208
	u32 last_lost;
2209
	int mib_idx;
2210
	int fwd_rexmitting = 0;
2211

2212 2213 2214
	if (!tp->lost_out)
		tp->retransmit_high = tp->snd_una;

2215
	if (tp->retransmit_skb_hint) {
2216
		skb = tp->retransmit_skb_hint;
2217 2218 2219 2220
		last_lost = TCP_SKB_CB(skb)->end_seq;
		if (after(last_lost, tp->retransmit_high))
			last_lost = tp->retransmit_high;
	} else {
2221
		skb = tcp_write_queue_head(sk);
2222 2223
		last_lost = tp->snd_una;
	}
L
Linus Torvalds 已提交
2224

2225 2226
	tcp_for_write_queue_from(skb, sk) {
		__u8 sacked = TCP_SKB_CB(skb)->sacked;
L
Linus Torvalds 已提交
2227

2228 2229 2230
		if (skb == tcp_send_head(sk))
			break;
		/* we could do better than to assign each time */
2231 2232
		if (hole == NULL)
			tp->retransmit_skb_hint = skb;
2233 2234 2235 2236 2237 2238 2239 2240 2241 2242

		/* Assume this retransmit will generate
		 * only one packet for congestion window
		 * calculation purposes.  This works because
		 * tcp_retransmit_skb() will chop up the
		 * packet to be MSS sized and all the
		 * packet counting works out.
		 */
		if (tcp_packets_in_flight(tp) >= tp->snd_cwnd)
			return;
L
Linus Torvalds 已提交
2243

2244 2245 2246 2247 2248
		if (fwd_rexmitting) {
begin_fwd:
			if (!before(TCP_SKB_CB(skb)->seq, tcp_highest_sack_seq(tp)))
				break;
			mib_idx = LINUX_MIB_TCPFORWARDRETRANS;
2249

2250
		} else if (!before(TCP_SKB_CB(skb)->seq, tp->retransmit_high)) {
2251
			tp->retransmit_high = last_lost;
2252 2253 2254 2255 2256 2257 2258 2259 2260
			if (!tcp_can_forward_retransmit(sk))
				break;
			/* Backtrack if necessary to non-L'ed skb */
			if (hole != NULL) {
				skb = hole;
				hole = NULL;
			}
			fwd_rexmitting = 1;
			goto begin_fwd;
L
Linus Torvalds 已提交
2261

2262
		} else if (!(sacked & TCPCB_LOST)) {
2263
			if (hole == NULL && !(sacked & (TCPCB_SACKED_RETRANS|TCPCB_SACKED_ACKED)))
2264 2265
				hole = skb;
			continue;
L
Linus Torvalds 已提交
2266

2267
		} else {
2268
			last_lost = TCP_SKB_CB(skb)->end_seq;
2269 2270 2271 2272 2273
			if (icsk->icsk_ca_state != TCP_CA_Loss)
				mib_idx = LINUX_MIB_TCPFASTRETRANS;
			else
				mib_idx = LINUX_MIB_TCPSLOWSTARTRETRANS;
		}
L
Linus Torvalds 已提交
2274

2275
		if (sacked & (TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))
L
Linus Torvalds 已提交
2276 2277
			continue;

2278
		if (tcp_retransmit_skb(sk, skb))
2279 2280
			return;
		NET_INC_STATS_BH(sock_net(sk), mib_idx);
L
Linus Torvalds 已提交
2281

2282
		if (skb == tcp_write_queue_head(sk))
2283 2284 2285
			inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
						  inet_csk(sk)->icsk_rto,
						  TCP_RTO_MAX);
L
Linus Torvalds 已提交
2286 2287 2288 2289 2290 2291 2292 2293
	}
}

/* Send a fin.  The caller locks the socket for us.  This cannot be
 * allowed to fail queueing a FIN frame under any circumstances.
 */
void tcp_send_fin(struct sock *sk)
{
2294
	struct tcp_sock *tp = tcp_sk(sk);
2295
	struct sk_buff *skb = tcp_write_queue_tail(sk);
L
Linus Torvalds 已提交
2296
	int mss_now;
2297

L
Linus Torvalds 已提交
2298 2299 2300 2301
	/* Optimization, tack on the FIN if we have a queue of
	 * unsent frames.  But be careful about outgoing SACKS
	 * and IP options.
	 */
I
Ilpo Järvinen 已提交
2302
	mss_now = tcp_current_mss(sk);
L
Linus Torvalds 已提交
2303

2304
	if (tcp_send_head(sk) != NULL) {
L
Linus Torvalds 已提交
2305 2306 2307 2308 2309 2310
		TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_FIN;
		TCP_SKB_CB(skb)->end_seq++;
		tp->write_seq++;
	} else {
		/* Socket is locked, keep trying until memory is available. */
		for (;;) {
2311 2312
			skb = alloc_skb_fclone(MAX_TCP_HEADER,
					       sk->sk_allocation);
L
Linus Torvalds 已提交
2313 2314 2315 2316 2317 2318 2319 2320
			if (skb)
				break;
			yield();
		}

		/* Reserve space for headers and prepare control bits. */
		skb_reserve(skb, MAX_TCP_HEADER);
		/* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */
2321 2322
		tcp_init_nondata_skb(skb, tp->write_seq,
				     TCPCB_FLAG_ACK | TCPCB_FLAG_FIN);
L
Linus Torvalds 已提交
2323 2324
		tcp_queue_skb(sk, skb);
	}
2325
	__tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_OFF);
L
Linus Torvalds 已提交
2326 2327 2328 2329 2330
}

/* We get here when a process closes a file descriptor (either due to
 * an explicit close() or as a byproduct of exit()'ing) and there
 * was unread data in the receive queue.  This behavior is recommended
2331
 * by RFC 2525, section 2.17.  -DaveM
L
Linus Torvalds 已提交
2332
 */
A
Al Viro 已提交
2333
void tcp_send_active_reset(struct sock *sk, gfp_t priority)
L
Linus Torvalds 已提交
2334 2335 2336 2337 2338 2339
{
	struct sk_buff *skb;

	/* NOTE: No TCP options attached and we never retransmit this. */
	skb = alloc_skb(MAX_TCP_HEADER, priority);
	if (!skb) {
P
Pavel Emelyanov 已提交
2340
		NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
L
Linus Torvalds 已提交
2341 2342 2343 2344 2345
		return;
	}

	/* Reserve space for headers and prepare control bits. */
	skb_reserve(skb, MAX_TCP_HEADER);
2346 2347
	tcp_init_nondata_skb(skb, tcp_acceptable_seq(sk),
			     TCPCB_FLAG_ACK | TCPCB_FLAG_RST);
L
Linus Torvalds 已提交
2348 2349
	/* Send it off. */
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2350
	if (tcp_transmit_skb(sk, skb, 0, priority))
P
Pavel Emelyanov 已提交
2351
		NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
2352

P
Pavel Emelyanov 已提交
2353
	TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTRSTS);
L
Linus Torvalds 已提交
2354 2355
}

2356 2357
/* Send a crossed SYN-ACK during socket establishment.
 * WARNING: This routine must only be called when we have already sent
L
Linus Torvalds 已提交
2358 2359 2360 2361 2362 2363
 * a SYN packet that crossed the incoming SYN that caused this routine
 * to get called. If this assumption fails then the initial rcv_wnd
 * and rcv_wscale values will not be correct.
 */
int tcp_send_synack(struct sock *sk)
{
2364
	struct sk_buff *skb;
L
Linus Torvalds 已提交
2365

2366
	skb = tcp_write_queue_head(sk);
2367
	if (skb == NULL || !(TCP_SKB_CB(skb)->flags & TCPCB_FLAG_SYN)) {
L
Linus Torvalds 已提交
2368 2369 2370
		printk(KERN_DEBUG "tcp_send_synack: wrong queue state\n");
		return -EFAULT;
	}
2371
	if (!(TCP_SKB_CB(skb)->flags & TCPCB_FLAG_ACK)) {
L
Linus Torvalds 已提交
2372 2373 2374 2375
		if (skb_cloned(skb)) {
			struct sk_buff *nskb = skb_copy(skb, GFP_ATOMIC);
			if (nskb == NULL)
				return -ENOMEM;
2376
			tcp_unlink_write_queue(skb, sk);
L
Linus Torvalds 已提交
2377
			skb_header_release(nskb);
2378
			__tcp_add_write_queue_head(sk, nskb);
2379 2380 2381
			sk_wmem_free_skb(sk, skb);
			sk->sk_wmem_queued += nskb->truesize;
			sk_mem_charge(sk, nskb->truesize);
L
Linus Torvalds 已提交
2382 2383 2384 2385 2386 2387 2388
			skb = nskb;
		}

		TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_ACK;
		TCP_ECN_send_synack(tcp_sk(sk), skb);
	}
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2389
	return tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
L
Linus Torvalds 已提交
2390 2391
}

2392
/* Prepare a SYN-ACK. */
2393
struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
2394 2395
				struct request_sock *req,
				struct request_values *rvp)
L
Linus Torvalds 已提交
2396
{
2397
	struct tcp_out_options opts;
2398
	struct tcp_extend_values *xvp = tcp_xv(rvp);
2399
	struct inet_request_sock *ireq = inet_rsk(req);
L
Linus Torvalds 已提交
2400 2401 2402
	struct tcp_sock *tp = tcp_sk(sk);
	struct tcphdr *th;
	struct sk_buff *skb;
2403
	struct tcp_md5sig_key *md5;
2404
	int tcp_header_size;
2405
	int mss;
L
Linus Torvalds 已提交
2406 2407 2408 2409 2410 2411 2412 2413

	skb = sock_wmalloc(sk, MAX_TCP_HEADER + 15, 1, GFP_ATOMIC);
	if (skb == NULL)
		return NULL;

	/* Reserve space for headers. */
	skb_reserve(skb, MAX_TCP_HEADER);

E
Eric Dumazet 已提交
2414
	skb_dst_set(skb, dst_clone(dst));
L
Linus Torvalds 已提交
2415

2416 2417 2418 2419
	mss = dst_metric(dst, RTAX_ADVMSS);
	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < mss)
		mss = tp->rx_opt.user_mss;

A
Adam Langley 已提交
2420 2421 2422 2423 2424 2425
	if (req->rcv_wnd == 0) { /* ignored for retransmitted syns */
		__u8 rcv_wscale;
		/* Set this up on the first call only */
		req->window_clamp = tp->window_clamp ? : dst_metric(dst, RTAX_WINDOW);
		/* tcp_full_space because it is guaranteed to be the first packet */
		tcp_select_initial_window(tcp_full_space(sk),
2426
			mss - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0),
A
Adam Langley 已提交
2427 2428 2429 2430 2431 2432 2433 2434
			&req->rcv_wnd,
			&req->window_clamp,
			ireq->wscale_ok,
			&rcv_wscale);
		ireq->rcv_wscale = rcv_wscale;
	}

	memset(&opts, 0, sizeof(opts));
2435 2436 2437 2438 2439
#ifdef CONFIG_SYN_COOKIES
	if (unlikely(req->cookie_ts))
		TCP_SKB_CB(skb)->when = cookie_init_timestamp(req);
	else
#endif
A
Adam Langley 已提交
2440
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2441
	tcp_header_size = tcp_synack_options(sk, req, mss,
2442 2443
					     skb, &opts, &md5, xvp)
			+ sizeof(*th);
2444

2445 2446
	skb_push(skb, tcp_header_size);
	skb_reset_transport_header(skb);
L
Linus Torvalds 已提交
2447

2448
	th = tcp_hdr(skb);
L
Linus Torvalds 已提交
2449 2450 2451 2452
	memset(th, 0, sizeof(struct tcphdr));
	th->syn = 1;
	th->ack = 1;
	TCP_ECN_make_synack(req, th);
2453
	th->source = ireq->loc_port;
2454
	th->dest = ireq->rmt_port;
2455 2456 2457 2458 2459
	/* Setting of flags are superfluous here for callers (and ECE is
	 * not even correctly set)
	 */
	tcp_init_nondata_skb(skb, tcp_rsk(req)->snt_isn,
			     TCPCB_FLAG_SYN | TCPCB_FLAG_ACK);
2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488

	if (OPTION_COOKIE_EXTENSION & opts.options) {
		const struct tcp_cookie_values *cvp = tp->cookie_values;

		if (cvp != NULL &&
		    cvp->s_data_constant &&
		    cvp->s_data_desired > 0) {
			u8 *buf = skb_put(skb, cvp->s_data_desired);

			/* copy data directly from the listening socket. */
			memcpy(buf, cvp->s_data_payload, cvp->s_data_desired);
			TCP_SKB_CB(skb)->end_seq += cvp->s_data_desired;
		}

		if (opts.hash_size > 0) {
			__u32 workspace[SHA_WORKSPACE_WORDS];
			u32 *mess = &xvp->cookie_bakery[COOKIE_DIGEST_WORDS];
			u32 *tail = &mess[COOKIE_MESSAGE_WORDS-1];

			/* Secret recipe depends on the Timestamp, (future)
			 * Sequence and Acknowledgment Numbers, Initiator
			 * Cookie, and others handled by IP variant caller.
			 */
			*tail-- ^= opts.tsval;
			*tail-- ^= tcp_rsk(req)->rcv_isn + 1;
			*tail-- ^= TCP_SKB_CB(skb)->seq + 1;

			/* recommended */
			*tail-- ^= ((th->dest << 16) | th->source);
D
David S. Miller 已提交
2489
			*tail-- ^= (u32)(unsigned long)cvp; /* per sockopt */
2490 2491 2492 2493 2494 2495 2496 2497 2498

			sha_transform((__u32 *)&xvp->cookie_bakery[0],
				      (char *)mess,
				      &workspace[0]);
			opts.hash_location =
				(__u8 *)&xvp->cookie_bakery[0];
		}
	}

L
Linus Torvalds 已提交
2499
	th->seq = htonl(TCP_SKB_CB(skb)->seq);
2500
	th->ack_seq = htonl(tcp_rsk(req)->rcv_isn + 1);
L
Linus Torvalds 已提交
2501 2502

	/* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */
2503
	th->window = htons(min(req->rcv_wnd, 65535U));
2504
	tcp_options_write((__be32 *)(th + 1), tp, &opts);
L
Linus Torvalds 已提交
2505
	th->doff = (tcp_header_size >> 2);
P
Pavel Emelyanov 已提交
2506
	TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTSEGS);
2507 2508 2509 2510

#ifdef CONFIG_TCP_MD5SIG
	/* Okay, we have all we need - do the md5 hash if needed */
	if (md5) {
2511
		tcp_rsk(req)->af_specific->calc_md5_hash(opts.hash_location,
2512
					       md5, NULL, req, skb);
2513 2514 2515
	}
#endif

L
Linus Torvalds 已提交
2516 2517 2518
	return skb;
}

2519
/* Do all connect socket setups that can be done AF independent. */
S
Stephen Hemminger 已提交
2520
static void tcp_connect_init(struct sock *sk)
L
Linus Torvalds 已提交
2521 2522 2523 2524 2525 2526 2527 2528 2529
{
	struct dst_entry *dst = __sk_dst_get(sk);
	struct tcp_sock *tp = tcp_sk(sk);
	__u8 rcv_wscale;

	/* We'll fix this up when we get a response from the other end.
	 * See tcp_input.c:tcp_rcv_state_process case TCP_SYN_SENT.
	 */
	tp->tcp_header_len = sizeof(struct tcphdr) +
2530 2531 2532
		(sysctl_tcp_timestamps &&
		(!dst_feature(dst, RTAX_FEATURE_NO_TSTAMP) ?
		  TCPOLEN_TSTAMP_ALIGNED : 0));
L
Linus Torvalds 已提交
2533

2534 2535 2536 2537 2538
#ifdef CONFIG_TCP_MD5SIG
	if (tp->af_specific->md5_lookup(sk, sk) != NULL)
		tp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED;
#endif

L
Linus Torvalds 已提交
2539 2540 2541 2542
	/* If user gave his TCP_MAXSEG, record it to clamp */
	if (tp->rx_opt.user_mss)
		tp->rx_opt.mss_clamp = tp->rx_opt.user_mss;
	tp->max_window = 0;
J
John Heffner 已提交
2543
	tcp_mtup_init(sk);
L
Linus Torvalds 已提交
2544 2545 2546 2547 2548
	tcp_sync_mss(sk, dst_mtu(dst));

	if (!tp->window_clamp)
		tp->window_clamp = dst_metric(dst, RTAX_WINDOW);
	tp->advmss = dst_metric(dst, RTAX_ADVMSS);
2549 2550 2551
	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < tp->advmss)
		tp->advmss = tp->rx_opt.user_mss;

L
Linus Torvalds 已提交
2552 2553 2554 2555 2556 2557
	tcp_initialize_rcv_mss(sk);

	tcp_select_initial_window(tcp_full_space(sk),
				  tp->advmss - (tp->rx_opt.ts_recent_stamp ? tp->tcp_header_len - sizeof(struct tcphdr) : 0),
				  &tp->rcv_wnd,
				  &tp->window_clamp,
2558 2559
				  (sysctl_tcp_window_scaling &&
				   !dst_feature(dst, RTAX_FEATURE_NO_WSCALE)),
L
Linus Torvalds 已提交
2560 2561 2562 2563 2564 2565 2566 2567
				  &rcv_wscale);

	tp->rx_opt.rcv_wscale = rcv_wscale;
	tp->rcv_ssthresh = tp->rcv_wnd;

	sk->sk_err = 0;
	sock_reset_flag(sk, SOCK_DONE);
	tp->snd_wnd = 0;
2568
	tcp_init_wl(tp, 0);
L
Linus Torvalds 已提交
2569 2570
	tp->snd_una = tp->write_seq;
	tp->snd_sml = tp->write_seq;
I
Ilpo Järvinen 已提交
2571
	tp->snd_up = tp->write_seq;
L
Linus Torvalds 已提交
2572 2573 2574 2575
	tp->rcv_nxt = 0;
	tp->rcv_wup = 0;
	tp->copied_seq = 0;

2576 2577
	inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT;
	inet_csk(sk)->icsk_retransmits = 0;
L
Linus Torvalds 已提交
2578 2579 2580
	tcp_clear_retrans(tp);
}

2581
/* Build a SYN and send it off. */
L
Linus Torvalds 已提交
2582 2583 2584 2585 2586 2587 2588
int tcp_connect(struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *buff;

	tcp_connect_init(sk);

2589
	buff = alloc_skb_fclone(MAX_TCP_HEADER + 15, sk->sk_allocation);
L
Linus Torvalds 已提交
2590 2591 2592 2593 2594 2595
	if (unlikely(buff == NULL))
		return -ENOBUFS;

	/* Reserve space for headers. */
	skb_reserve(buff, MAX_TCP_HEADER);

2596
	tp->snd_nxt = tp->write_seq;
2597 2598
	tcp_init_nondata_skb(buff, tp->write_seq++, TCPCB_FLAG_SYN);
	TCP_ECN_send_syn(sk, buff);
L
Linus Torvalds 已提交
2599 2600 2601 2602 2603

	/* Send it off. */
	TCP_SKB_CB(buff)->when = tcp_time_stamp;
	tp->retrans_stamp = TCP_SKB_CB(buff)->when;
	skb_header_release(buff);
2604
	__tcp_add_write_queue_tail(sk, buff);
2605 2606
	sk->sk_wmem_queued += buff->truesize;
	sk_mem_charge(sk, buff->truesize);
L
Linus Torvalds 已提交
2607
	tp->packets_out += tcp_skb_pcount(buff);
2608
	tcp_transmit_skb(sk, buff, 1, sk->sk_allocation);
2609 2610 2611 2612 2613 2614

	/* We change tp->snd_nxt after the tcp_transmit_skb() call
	 * in order to make this packet get counted in tcpOutSegs.
	 */
	tp->snd_nxt = tp->write_seq;
	tp->pushed_seq = tp->write_seq;
P
Pavel Emelyanov 已提交
2615
	TCP_INC_STATS(sock_net(sk), TCP_MIB_ACTIVEOPENS);
L
Linus Torvalds 已提交
2616 2617

	/* Timer for repeating the SYN until an answer. */
2618 2619
	inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
				  inet_csk(sk)->icsk_rto, TCP_RTO_MAX);
L
Linus Torvalds 已提交
2620 2621 2622 2623 2624 2625 2626 2627 2628
	return 0;
}

/* Send out a delayed ack, the caller does the policy checking
 * to see if we should even be here.  See tcp_input.c:tcp_ack_snd_check()
 * for details.
 */
void tcp_send_delayed_ack(struct sock *sk)
{
2629 2630
	struct inet_connection_sock *icsk = inet_csk(sk);
	int ato = icsk->icsk_ack.ato;
L
Linus Torvalds 已提交
2631 2632 2633
	unsigned long timeout;

	if (ato > TCP_DELACK_MIN) {
2634
		const struct tcp_sock *tp = tcp_sk(sk);
2635
		int max_ato = HZ / 2;
L
Linus Torvalds 已提交
2636

2637 2638
		if (icsk->icsk_ack.pingpong ||
		    (icsk->icsk_ack.pending & ICSK_ACK_PUSHED))
L
Linus Torvalds 已提交
2639 2640 2641 2642 2643
			max_ato = TCP_DELACK_MAX;

		/* Slow path, intersegment interval is "high". */

		/* If some rtt estimate is known, use it to bound delayed ack.
2644
		 * Do not use inet_csk(sk)->icsk_rto here, use results of rtt measurements
L
Linus Torvalds 已提交
2645 2646 2647
		 * directly.
		 */
		if (tp->srtt) {
2648
			int rtt = max(tp->srtt >> 3, TCP_DELACK_MIN);
L
Linus Torvalds 已提交
2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660

			if (rtt < max_ato)
				max_ato = rtt;
		}

		ato = min(ato, max_ato);
	}

	/* Stay within the limit we were given */
	timeout = jiffies + ato;

	/* Use new timeout only if there wasn't a older one earlier. */
2661
	if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) {
L
Linus Torvalds 已提交
2662 2663 2664
		/* If delack timer was blocked or is about to expire,
		 * send ACK now.
		 */
2665 2666
		if (icsk->icsk_ack.blocked ||
		    time_before_eq(icsk->icsk_ack.timeout, jiffies + (ato >> 2))) {
L
Linus Torvalds 已提交
2667 2668 2669 2670
			tcp_send_ack(sk);
			return;
		}

2671 2672
		if (!time_before(timeout, icsk->icsk_ack.timeout))
			timeout = icsk->icsk_ack.timeout;
L
Linus Torvalds 已提交
2673
	}
2674 2675 2676
	icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER;
	icsk->icsk_ack.timeout = timeout;
	sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout);
L
Linus Torvalds 已提交
2677 2678 2679 2680 2681
}

/* This routine sends an ack and also updates the window. */
void tcp_send_ack(struct sock *sk)
{
2682
	struct sk_buff *buff;
L
Linus Torvalds 已提交
2683

2684 2685 2686
	/* If we have been reset, we may not send again. */
	if (sk->sk_state == TCP_CLOSE)
		return;
L
Linus Torvalds 已提交
2687

2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698
	/* We are not putting this on the write queue, so
	 * tcp_transmit_skb() will set the ownership to this
	 * sock.
	 */
	buff = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC);
	if (buff == NULL) {
		inet_csk_schedule_ack(sk);
		inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN;
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
					  TCP_DELACK_MAX, TCP_RTO_MAX);
		return;
L
Linus Torvalds 已提交
2699
	}
2700 2701 2702

	/* Reserve space for headers and prepare control bits. */
	skb_reserve(buff, MAX_TCP_HEADER);
2703
	tcp_init_nondata_skb(buff, tcp_acceptable_seq(sk), TCPCB_FLAG_ACK);
2704 2705 2706 2707

	/* Send it off, this clears delayed acks for us. */
	TCP_SKB_CB(buff)->when = tcp_time_stamp;
	tcp_transmit_skb(sk, buff, 0, GFP_ATOMIC);
L
Linus Torvalds 已提交
2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727
}

/* This routine sends a packet with an out of date sequence
 * number. It assumes the other end will try to ack it.
 *
 * Question: what should we make while urgent mode?
 * 4.4BSD forces sending single byte of data. We cannot send
 * out of window data, because we have SND.NXT==SND.MAX...
 *
 * Current solution: to send TWO zero-length segments in urgent mode:
 * one is with SEG.SEQ=SND.UNA to deliver urgent pointer, another is
 * out-of-date with SND.UNA-1 to probe window.
 */
static int tcp_xmit_probe_skb(struct sock *sk, int urgent)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;

	/* We don't queue it, tcp_transmit_skb() sets ownership. */
	skb = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC);
2728
	if (skb == NULL)
L
Linus Torvalds 已提交
2729 2730 2731 2732 2733 2734 2735 2736
		return -1;

	/* Reserve space for headers and set control bits. */
	skb_reserve(skb, MAX_TCP_HEADER);
	/* Use a previous sequence.  This should cause the other
	 * end to send an ack.  Don't queue or clone SKB, just
	 * send it.
	 */
2737
	tcp_init_nondata_skb(skb, tp->snd_una - !urgent, TCPCB_FLAG_ACK);
L
Linus Torvalds 已提交
2738
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2739
	return tcp_transmit_skb(sk, skb, 0, GFP_ATOMIC);
L
Linus Torvalds 已提交
2740 2741
}

2742
/* Initiate keepalive or window probe from timer. */
L
Linus Torvalds 已提交
2743 2744
int tcp_write_wakeup(struct sock *sk)
{
2745 2746
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;
L
Linus Torvalds 已提交
2747

2748 2749 2750 2751 2752 2753
	if (sk->sk_state == TCP_CLOSE)
		return -1;

	if ((skb = tcp_send_head(sk)) != NULL &&
	    before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp))) {
		int err;
I
Ilpo Järvinen 已提交
2754
		unsigned int mss = tcp_current_mss(sk);
2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766
		unsigned int seg_size = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;

		if (before(tp->pushed_seq, TCP_SKB_CB(skb)->end_seq))
			tp->pushed_seq = TCP_SKB_CB(skb)->end_seq;

		/* We are probing the opening of a window
		 * but the window size is != 0
		 * must have been a result SWS avoidance ( sender )
		 */
		if (seg_size < TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq ||
		    skb->len > mss) {
			seg_size = min(seg_size, mss);
L
Linus Torvalds 已提交
2767
			TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779
			if (tcp_fragment(sk, skb, seg_size, mss))
				return -1;
		} else if (!tcp_skb_pcount(skb))
			tcp_set_skb_tso_segs(sk, skb, mss);

		TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
		TCP_SKB_CB(skb)->when = tcp_time_stamp;
		err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
		if (!err)
			tcp_event_new_data_sent(sk, skb);
		return err;
	} else {
I
Ilpo Järvinen 已提交
2780
		if (between(tp->snd_up, tp->snd_una + 1, tp->snd_una + 0xFFFF))
2781 2782
			tcp_xmit_probe_skb(sk, 1);
		return tcp_xmit_probe_skb(sk, 0);
L
Linus Torvalds 已提交
2783 2784 2785 2786 2787 2788 2789 2790
	}
}

/* A window probe timeout has occurred.  If window is not closed send
 * a partial packet else a zero probe.
 */
void tcp_send_probe0(struct sock *sk)
{
2791
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
2792 2793 2794 2795 2796
	struct tcp_sock *tp = tcp_sk(sk);
	int err;

	err = tcp_write_wakeup(sk);

2797
	if (tp->packets_out || !tcp_send_head(sk)) {
L
Linus Torvalds 已提交
2798
		/* Cancel probe timer, if it is not required. */
2799
		icsk->icsk_probes_out = 0;
2800
		icsk->icsk_backoff = 0;
L
Linus Torvalds 已提交
2801 2802 2803 2804
		return;
	}

	if (err <= 0) {
2805 2806
		if (icsk->icsk_backoff < sysctl_tcp_retries2)
			icsk->icsk_backoff++;
2807
		icsk->icsk_probes_out++;
2808
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
2809 2810
					  min(icsk->icsk_rto << icsk->icsk_backoff, TCP_RTO_MAX),
					  TCP_RTO_MAX);
L
Linus Torvalds 已提交
2811 2812
	} else {
		/* If packet was not sent due to local congestion,
2813
		 * do not backoff and do not remember icsk_probes_out.
L
Linus Torvalds 已提交
2814 2815 2816 2817
		 * Let local senders to fight for local resources.
		 *
		 * Use accumulated backoff yet.
		 */
2818 2819
		if (!icsk->icsk_probes_out)
			icsk->icsk_probes_out = 1;
2820
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
2821
					  min(icsk->icsk_rto << icsk->icsk_backoff,
2822 2823
					      TCP_RESOURCE_PROBE_INTERVAL),
					  TCP_RTO_MAX);
L
Linus Torvalds 已提交
2824 2825 2826
	}
}

2827
EXPORT_SYMBOL(tcp_select_initial_window);
L
Linus Torvalds 已提交
2828 2829 2830 2831
EXPORT_SYMBOL(tcp_connect);
EXPORT_SYMBOL(tcp_make_synack);
EXPORT_SYMBOL(tcp_simple_retransmit);
EXPORT_SYMBOL(tcp_sync_mss);
J
John Heffner 已提交
2832
EXPORT_SYMBOL(tcp_mtup_init);