tcp_output.c 68.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * 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).
 *
 * Version:	$Id: tcp_output.c,v 1.146 2002/02/01 22:01:04 davem Exp $
 *
10
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
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 43 44 45 46 47
 *		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>
#include <linux/smp_lock.h>

/* People can turn this off for buggy TCP's found in printers etc. */
int sysctl_tcp_retrans_collapse = 1;

48 49 50 51 52
/* People can turn this on to  work with those rare, broken TCPs that
 * interpret the window field as a signed quantity.
 */
int sysctl_tcp_workaround_signed_windows = 0;

L
Linus Torvalds 已提交
53 54 55 56
/* 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.
 */
57
int sysctl_tcp_tso_win_divisor = 3;
L
Linus Torvalds 已提交
58

J
John Heffner 已提交
59 60 61
int sysctl_tcp_mtu_probing = 0;
int sysctl_tcp_base_mss = 512;

S
Stephen Hemminger 已提交
62 63
static void update_send_head(struct sock *sk, struct tcp_sock *tp,
			     struct sk_buff *skb)
L
Linus Torvalds 已提交
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
{
	sk->sk_send_head = skb->next;
	if (sk->sk_send_head == (struct sk_buff *)&sk->sk_write_queue)
		sk->sk_send_head = NULL;
	tp->snd_nxt = TCP_SKB_CB(skb)->end_seq;
	tcp_packets_out_inc(sk, tp, skb);
}

/* 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:
 */
static inline __u32 tcp_acceptable_seq(struct sock *sk, struct tcp_sock *tp)
{
	if (!before(tp->snd_una+tp->snd_wnd, tp->snd_nxt))
		return tp->snd_nxt;
	else
		return tp->snd_una+tp->snd_wnd;
}

/* 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. */
116
static void tcp_cwnd_restart(struct sock *sk, struct dst_entry *dst)
L
Linus Torvalds 已提交
117
{
118
	struct tcp_sock *tp = tcp_sk(sk);
L
Linus Torvalds 已提交
119 120 121 122
	s32 delta = tcp_time_stamp - tp->lsndtime;
	u32 restart_cwnd = tcp_init_cwnd(tp, dst);
	u32 cwnd = tp->snd_cwnd;

123
	tcp_ca_event(sk, CA_EVENT_CWND_RESTART);
L
Linus Torvalds 已提交
124

125
	tp->snd_ssthresh = tcp_current_ssthresh(sk);
L
Linus Torvalds 已提交
126 127
	restart_cwnd = min(restart_cwnd, cwnd);

128
	while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd)
L
Linus Torvalds 已提交
129 130 131 132 133 134
		cwnd >>= 1;
	tp->snd_cwnd = max(cwnd, restart_cwnd);
	tp->snd_cwnd_stamp = tcp_time_stamp;
	tp->snd_cwnd_used = 0;
}

S
Stephen Hemminger 已提交
135 136
static void tcp_event_data_sent(struct tcp_sock *tp,
				struct sk_buff *skb, struct sock *sk)
L
Linus Torvalds 已提交
137
{
138 139
	struct inet_connection_sock *icsk = inet_csk(sk);
	const u32 now = tcp_time_stamp;
L
Linus Torvalds 已提交
140

141 142
	if (!tp->packets_out && (s32)(now - tp->lsndtime) > icsk->icsk_rto)
		tcp_cwnd_restart(sk, __sk_dst_get(sk));
L
Linus Torvalds 已提交
143 144 145 146 147 148

	tp->lsndtime = now;

	/* If it is a reply for ato after last received
	 * packet, enter pingpong mode.
	 */
149 150
	if ((u32)(now - icsk->icsk_ack.lrcvtime) < icsk->icsk_ack.ato)
		icsk->icsk_ack.pingpong = 1;
L
Linus Torvalds 已提交
151 152
}

S
Stephen Hemminger 已提交
153
static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts)
L
Linus Torvalds 已提交
154
{
155 156
	tcp_dec_quickack_mode(sk, pkts);
	inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
L
Linus Torvalds 已提交
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
}

/* 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
182 183 184 185 186 187
	 * 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
188
	 */
189 190 191 192 193
	if (sysctl_tcp_workaround_signed_windows)
		(*rcv_wnd) = min(space, MAX_TCP_WINDOW);
	else
		(*rcv_wnd) = space;

L
Linus Torvalds 已提交
194 195 196 197 198 199 200 201 202 203 204 205 206
	(*rcv_wscale) = 0;
	if (wscale_ok) {
		/* Set window scaling on max possible window
		 * See RFC1323 for an explanation of the limit to 14 
		 */
		space = max_t(u32, sysctl_tcp_rmem[2], sysctl_rmem_max);
		while (space > 65535 && (*rcv_wscale) < 14) {
			space >>= 1;
			(*rcv_wscale)++;
		}
	}

	/* Set initial window to value enough for senders,
207
	 * following RFC2414. Senders, not following this RFC,
L
Linus Torvalds 已提交
208 209 210
	 * will be satisfied with 2.
	 */
	if (mss > (1<<*rcv_wscale)) {
211 212
		int init_cwnd = 4;
		if (mss > 1460*3)
L
Linus Torvalds 已提交
213
			init_cwnd = 2;
214 215
		else if (mss > 1460)
			init_cwnd = 3;
L
Linus Torvalds 已提交
216 217 218 219 220 221 222 223 224 225 226 227 228
		if (*rcv_wnd > init_cwnd*mss)
			*rcv_wnd = init_cwnd*mss;
	}

	/* 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
229
static u16 tcp_select_window(struct sock *sk)
L
Linus Torvalds 已提交
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251
{
	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 */
	if(new_win < cur_win) {
		/* 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.
		 */
		new_win = cur_win;
	}
	tp->rcv_wnd = new_win;
	tp->rcv_wup = tp->rcv_nxt;

	/* Make sure we do not exceed the maximum possible
	 * scaled window.
	 */
252
	if (!tp->rx_opt.rcv_wscale && sysctl_tcp_workaround_signed_windows)
L
Linus Torvalds 已提交
253 254 255 256 257 258 259 260 261 262 263 264 265 266
		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;
}

S
Stephen Hemminger 已提交
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335
static void tcp_build_and_update_options(__u32 *ptr, struct tcp_sock *tp,
					 __u32 tstamp)
{
	if (tp->rx_opt.tstamp_ok) {
		*ptr++ = __constant_htonl((TCPOPT_NOP << 24) |
					  (TCPOPT_NOP << 16) |
					  (TCPOPT_TIMESTAMP << 8) |
					  TCPOLEN_TIMESTAMP);
		*ptr++ = htonl(tstamp);
		*ptr++ = htonl(tp->rx_opt.ts_recent);
	}
	if (tp->rx_opt.eff_sacks) {
		struct tcp_sack_block *sp = tp->rx_opt.dsack ? tp->duplicate_sack : tp->selective_acks;
		int this_sack;

		*ptr++ = htonl((TCPOPT_NOP  << 24) |
			       (TCPOPT_NOP  << 16) |
			       (TCPOPT_SACK <<  8) |
			       (TCPOLEN_SACK_BASE + (tp->rx_opt.eff_sacks *
						     TCPOLEN_SACK_PERBLOCK)));
		for(this_sack = 0; this_sack < tp->rx_opt.eff_sacks; this_sack++) {
			*ptr++ = htonl(sp[this_sack].start_seq);
			*ptr++ = htonl(sp[this_sack].end_seq);
		}
		if (tp->rx_opt.dsack) {
			tp->rx_opt.dsack = 0;
			tp->rx_opt.eff_sacks--;
		}
	}
}

/* Construct a tcp options header for a SYN or SYN_ACK packet.
 * If this is every changed make sure to change the definition of
 * MAX_SYN_SIZE to match the new maximum number of options that you
 * can generate.
 */
static void tcp_syn_build_options(__u32 *ptr, int mss, int ts, int sack,
				  int offer_wscale, int wscale, __u32 tstamp,
				  __u32 ts_recent)
{
	/* 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.
	 */
	*ptr++ = htonl((TCPOPT_MSS << 24) | (TCPOLEN_MSS << 16) | mss);
	if (ts) {
		if(sack)
			*ptr++ = __constant_htonl((TCPOPT_SACK_PERM << 24) | (TCPOLEN_SACK_PERM << 16) |
						  (TCPOPT_TIMESTAMP << 8) | TCPOLEN_TIMESTAMP);
		else
			*ptr++ = __constant_htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
						  (TCPOPT_TIMESTAMP << 8) | TCPOLEN_TIMESTAMP);
		*ptr++ = htonl(tstamp);		/* TSVAL */
		*ptr++ = htonl(ts_recent);	/* TSECR */
	} else if(sack)
		*ptr++ = __constant_htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
					  (TCPOPT_SACK_PERM << 8) | TCPOLEN_SACK_PERM);
	if (offer_wscale)
		*ptr++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_WINDOW << 16) | (TCPOLEN_WINDOW << 8) | (wscale));
}
L
Linus Torvalds 已提交
336 337 338 339 340 341 342 343 344 345 346 347

/* 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.
 */
348
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
349
{
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
	const struct inet_connection_sock *icsk = inet_csk(sk);
	struct inet_sock *inet;
	struct tcp_sock *tp;
	struct tcp_skb_cb *tcb;
	int tcp_header_size;
	struct tcphdr *th;
	int sysctl_flags;
	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.
	 */
	if (icsk->icsk_ca_ops->rtt_sample)
		__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
375

376 377 378 379
	inet = inet_sk(sk);
	tp = tcp_sk(sk);
	tcb = TCP_SKB_CB(skb);
	tcp_header_size = tp->tcp_header_len;
L
Linus Torvalds 已提交
380 381 382 383 384

#define SYSCTL_FLAG_TSTAMPS	0x1
#define SYSCTL_FLAG_WSCALE	0x2
#define SYSCTL_FLAG_SACK	0x4

385 386 387 388 389 390
	sysctl_flags = 0;
	if (unlikely(tcb->flags & TCPCB_FLAG_SYN)) {
		tcp_header_size = sizeof(struct tcphdr) + TCPOLEN_MSS;
		if(sysctl_tcp_timestamps) {
			tcp_header_size += TCPOLEN_TSTAMP_ALIGNED;
			sysctl_flags |= SYSCTL_FLAG_TSTAMPS;
L
Linus Torvalds 已提交
391
		}
392 393 394
		if (sysctl_tcp_window_scaling) {
			tcp_header_size += TCPOLEN_WSCALE_ALIGNED;
			sysctl_flags |= SYSCTL_FLAG_WSCALE;
L
Linus Torvalds 已提交
395
		}
396 397 398 399
		if (sysctl_tcp_sack) {
			sysctl_flags |= SYSCTL_FLAG_SACK;
			if (!(sysctl_flags & SYSCTL_FLAG_TSTAMPS))
				tcp_header_size += TCPOLEN_SACKPERM_ALIGNED;
L
Linus Torvalds 已提交
400
		}
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
	} else if (unlikely(tp->rx_opt.eff_sacks)) {
		/* A SACK is 2 pad bytes, a 2 byte header, plus
		 * 2 32-bit sequence numbers for each SACK block.
		 */
		tcp_header_size += (TCPOLEN_SACK_BASE_ALIGNED +
				    (tp->rx_opt.eff_sacks *
				     TCPOLEN_SACK_PERBLOCK));
	}
		
	if (tcp_packets_in_flight(tp) == 0)
		tcp_ca_event(sk, CA_EVENT_TX_START);

	th = (struct tcphdr *) skb_push(skb, tcp_header_size);
	skb->h.th = th;
	skb_set_owner_w(skb, sk);

	/* Build TCP header and checksum it. */
	th->source		= inet->sport;
	th->dest		= inet->dport;
	th->seq			= htonl(tcb->seq);
	th->ack_seq		= htonl(tp->rcv_nxt);
	*(((__u16 *)th) + 6)	= htons(((tcp_header_size >> 2) << 12) |
					tcb->flags);

	if (unlikely(tcb->flags & TCPCB_FLAG_SYN)) {
		/* RFC1323: The window in SYN & SYN/ACK segments
		 * is never scaled.
		 */
		th->window	= htons(tp->rcv_wnd);
	} else {
		th->window	= htons(tcp_select_window(sk));
	}
	th->check		= 0;
	th->urg_ptr		= 0;
L
Linus Torvalds 已提交
435

436 437 438 439 440
	if (unlikely(tp->urg_mode &&
		     between(tp->snd_up, tcb->seq+1, tcb->seq+0xFFFF))) {
		th->urg_ptr		= htons(tp->snd_up-tcb->seq);
		th->urg			= 1;
	}
L
Linus Torvalds 已提交
441

442 443 444 445 446 447 448 449 450 451 452 453 454 455
	if (unlikely(tcb->flags & TCPCB_FLAG_SYN)) {
		tcp_syn_build_options((__u32 *)(th + 1),
				      tcp_advertise_mss(sk),
				      (sysctl_flags & SYSCTL_FLAG_TSTAMPS),
				      (sysctl_flags & SYSCTL_FLAG_SACK),
				      (sysctl_flags & SYSCTL_FLAG_WSCALE),
				      tp->rx_opt.rcv_wscale,
				      tcb->when,
				      tp->rx_opt.ts_recent);
	} else {
		tcp_build_and_update_options((__u32 *)(th + 1),
					     tp, tcb->when);
		TCP_ECN_send(sk, tp, skb, tcp_header_size);
	}
L
Linus Torvalds 已提交
456

457
	icsk->icsk_af_ops->send_check(sk, skb->len, skb);
L
Linus Torvalds 已提交
458

459 460
	if (likely(tcb->flags & TCPCB_FLAG_ACK))
		tcp_event_ack_sent(sk, tcp_skb_pcount(skb));
L
Linus Torvalds 已提交
461

462 463
	if (skb->len != tcp_header_size)
		tcp_event_data_sent(tp, skb, sk);
L
Linus Torvalds 已提交
464

465
	TCP_INC_STATS(TCP_MIB_OUTSEGS);
L
Linus Torvalds 已提交
466

467
	err = icsk->icsk_af_ops->queue_xmit(skb, 0);
468 469 470 471 472 473 474 475 476 477 478 479
	if (unlikely(err <= 0))
		return err;

	tcp_enter_cwr(sk);

	/* NET_XMIT_CN is special. It does not guarantee,
	 * that this packet is lost. It tells that device
	 * is about to start to drop packets or already
	 * drops some packets of the same priority and
	 * invokes us to send less aggressively.
	 */
	return err == NET_XMIT_CN ? 0 : err;
L
Linus Torvalds 已提交
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

#undef SYSCTL_FLAG_TSTAMPS
#undef SYSCTL_FLAG_WSCALE
#undef SYSCTL_FLAG_SACK
}


/* This routine just queue's the buffer 
 *
 * 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);
	__skb_queue_tail(&sk->sk_write_queue, skb);
	sk_charge_skb(sk, skb);

	/* Queue it, remembering where we must start sending. */
	if (sk->sk_send_head == NULL)
		sk->sk_send_head = skb;
}

507
static void tcp_set_skb_tso_segs(struct sock *sk, struct sk_buff *skb, unsigned int mss_now)
508
{
509
	if (skb->len <= mss_now ||
510 511 512 513 514 515 516 517 518
	    !(sk->sk_route_caps & NETIF_F_TSO)) {
		/* Avoid the costly divide in the normal
		 * non-TSO case.
		 */
		skb_shinfo(skb)->tso_segs = 1;
		skb_shinfo(skb)->tso_size = 0;
	} else {
		unsigned int factor;

519 520
		factor = skb->len + (mss_now - 1);
		factor /= mss_now;
521
		skb_shinfo(skb)->tso_segs = factor;
522
		skb_shinfo(skb)->tso_size = mss_now;
L
Linus Torvalds 已提交
523 524 525 526 527 528 529 530
	}
}

/* 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
 * packet to the list.  This won't be called frequently, I hope. 
 * Remember, these are still headerless SKBs at this point.
 */
531
int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss_now)
L
Linus Torvalds 已提交
532 533 534
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *buff;
535
	int nsize, old_factor;
L
Linus Torvalds 已提交
536 537
	u16 flags;

538
	BUG_ON(len > skb->len);
539 540

 	clear_all_retrans_hints(tp);
L
Linus Torvalds 已提交
541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564
	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. */
	sk_charge_skb(sk, buff);

	/* 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;
	TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
	TCP_SKB_CB(buff)->flags = flags;
565
	TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked;
L
Linus Torvalds 已提交
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
	TCP_SKB_CB(skb)->sacked &= ~TCPCB_AT_TAIL;

	if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_HW) {
		/* Copy and checksum data tail into the new buffer. */
		buff->csum = csum_partial_copy_nocheck(skb->data + len, skb_put(buff, nsize),
						       nsize, 0);

		skb_trim(skb, len);

		skb->csum = csum_block_sub(skb->csum, buff->csum, len);
	} else {
		skb->ip_summed = CHECKSUM_HW;
		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;
587
	buff->tstamp = skb->tstamp;
L
Linus Torvalds 已提交
588

589 590
	old_factor = tcp_skb_pcount(skb);

L
Linus Torvalds 已提交
591
	/* Fix up tso_factor for both original and new SKB.  */
592 593
	tcp_set_skb_tso_segs(sk, skb, mss_now);
	tcp_set_skb_tso_segs(sk, buff, mss_now);
L
Linus Torvalds 已提交
594

595 596 597
	/* If this packet has been sent out already, we must
	 * adjust the various packet counters.
	 */
598
	if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq)) {
599 600
		int diff = old_factor - tcp_skb_pcount(skb) -
			tcp_skb_pcount(buff);
L
Linus Torvalds 已提交
601

602
		tp->packets_out -= diff;
603 604 605 606 607 608

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

609 610 611 612
		if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) {
			tp->lost_out -= diff;
			tp->left_out -= diff;
		}
613

614
		if (diff > 0) {
615 616 617 618 619 620 621 622
			/* Adjust Reno SACK estimate. */
			if (!tp->rx_opt.sack_ok) {
				tp->sacked_out -= diff;
				if ((int)tp->sacked_out < 0)
					tp->sacked_out = 0;
				tcp_sync_left_out(tp);
			}

623 624 625 626
			tp->fackets_out -= diff;
			if ((int)tp->fackets_out < 0)
				tp->fackets_out = 0;
		}
L
Linus Torvalds 已提交
627 628 629
	}

	/* Link BUFF into the send queue. */
630
	skb_header_release(buff);
D
David S. Miller 已提交
631
	__skb_append(skb, buff, &sk->sk_write_queue);
L
Linus Torvalds 已提交
632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692

	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.
 */
static unsigned char *__pskb_trim_head(struct sk_buff *skb, int len)
{
	int i, k, eat;

	eat = len;
	k = 0;
	for (i=0; i<skb_shinfo(skb)->nr_frags; i++) {
		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;

	skb->tail = skb->data;
	skb->data_len -= len;
	skb->len = skb->data_len;
	return skb->tail;
}

int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len)
{
	if (skb_cloned(skb) &&
	    pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
		return -ENOMEM;

	if (len <= skb_headlen(skb)) {
		__skb_pull(skb, len);
	} else {
		if (__pskb_trim_head(skb, len-skb_headlen(skb)) == NULL)
			return -ENOMEM;
	}

	TCP_SKB_CB(skb)->seq += len;
	skb->ip_summed = CHECKSUM_HW;

	skb->truesize	     -= len;
	sk->sk_wmem_queued   -= len;
	sk->sk_forward_alloc += len;
	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)
693
		tcp_set_skb_tso_segs(sk, skb, tcp_current_mss(sk, 1));
L
Linus Torvalds 已提交
694 695 696 697

	return 0;
}

J
John Heffner 已提交
698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
/* Not accounting for SACKs here. */
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;
}

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) +
	                       icsk->icsk_af_ops->net_header_len;
	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
754 755 756 757 758 759
/* 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
760
   It is minimum of user_mss and mss received with SYN.
L
Linus Torvalds 已提交
761 762
   It also does not include TCP options.

763
   inet_csk(sk)->icsk_pmtu_cookie is last pmtu, seen by this function.
L
Linus Torvalds 已提交
764 765 766 767 768 769 770 771 772

   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.

773 774
   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
775 776 777 778 779
 */

unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
{
	struct tcp_sock *tp = tcp_sk(sk);
780
	struct inet_connection_sock *icsk = inet_csk(sk);
J
John Heffner 已提交
781
	int mss_now;
L
Linus Torvalds 已提交
782

J
John Heffner 已提交
783 784
	if (icsk->icsk_mtup.search_high > pmtu)
		icsk->icsk_mtup.search_high = pmtu;
L
Linus Torvalds 已提交
785

J
John Heffner 已提交
786
	mss_now = tcp_mtu_to_mss(sk, pmtu);
L
Linus Torvalds 已提交
787 788 789 790 791 792

	/* Bound mss with half of window */
	if (tp->max_window && mss_now > (tp->max_window>>1))
		mss_now = max((tp->max_window>>1), 68U - tp->tcp_header_len);

	/* And store cached results */
793
	icsk->icsk_pmtu_cookie = pmtu;
J
John Heffner 已提交
794 795
	if (icsk->icsk_mtup.enabled)
		mss_now = min(mss_now, tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_low));
796
	tp->mss_cache = mss_now;
L
Linus Torvalds 已提交
797 798 799 800 801 802 803 804 805 806 807

	return mss_now;
}

/* Compute the current effective MSS, taking SACKs and IP options,
 * and even PMTU discovery events into account.
 *
 * LARGESEND note: !urg_mode is overkill, only frames up to snd_up
 * cannot be large. However, taking into account rare use of URG, this
 * is not a big flaw.
 */
808
unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
L
Linus Torvalds 已提交
809 810 811
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct dst_entry *dst = __sk_dst_get(sk);
812 813 814 815 816 817 818 819 820 821
	u32 mss_now;
	u16 xmit_size_goal;
	int doing_tso = 0;

	mss_now = tp->mss_cache;

	if (large_allowed &&
	    (sk->sk_route_caps & NETIF_F_TSO) &&
	    !tp->urg_mode)
		doing_tso = 1;
L
Linus Torvalds 已提交
822 823 824

	if (dst) {
		u32 mtu = dst_mtu(dst);
825
		if (mtu != inet_csk(sk)->icsk_pmtu_cookie)
L
Linus Torvalds 已提交
826 827 828
			mss_now = tcp_sync_mss(sk, mtu);
	}

829 830 831
	if (tp->rx_opt.eff_sacks)
		mss_now -= (TCPOLEN_SACK_BASE_ALIGNED +
			    (tp->rx_opt.eff_sacks * TCPOLEN_SACK_PERBLOCK));
L
Linus Torvalds 已提交
832

833
	xmit_size_goal = mss_now;
L
Linus Torvalds 已提交
834

835
	if (doing_tso) {
836 837
		xmit_size_goal = (65535 -
				  inet_csk(sk)->icsk_af_ops->net_header_len -
838 839
				  inet_csk(sk)->icsk_ext_hdr_len -
				  tp->tcp_header_len);
L
Linus Torvalds 已提交
840

841 842 843 844
		if (tp->max_window &&
		    (xmit_size_goal > (tp->max_window >> 1)))
			xmit_size_goal = max((tp->max_window >> 1),
					     68U - tp->tcp_header_len);
L
Linus Torvalds 已提交
845

846
		xmit_size_goal -= (xmit_size_goal % mss_now);
L
Linus Torvalds 已提交
847
	}
848
	tp->xmit_size_goal = xmit_size_goal;
L
Linus Torvalds 已提交
849 850 851 852

	return mss_now;
}

853 854
/* Congestion window validation. (RFC2861) */

S
Stephen Hemminger 已提交
855
static void tcp_cwnd_validate(struct sock *sk, struct tcp_sock *tp)
856 857 858 859 860 861 862 863 864 865 866 867
{
	__u32 packets_out = tp->packets_out;

	if (packets_out >= tp->snd_cwnd) {
		/* 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;

868
		if ((s32)(tcp_time_stamp - tp->snd_cwnd_stamp) >= inet_csk(sk)->icsk_rto)
869 870 871 872
			tcp_cwnd_application_limited(sk);
	}
}

873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903
static unsigned int tcp_window_allows(struct tcp_sock *tp, struct sk_buff *skb, unsigned int mss_now, unsigned int cwnd)
{
	u32 window, cwnd_len;

	window = (tp->snd_una + tp->snd_wnd - TCP_SKB_CB(skb)->seq);
	cwnd_len = mss_now * cwnd;
	return min(window, cwnd_len);
}

/* 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.
 */
static inline unsigned int tcp_cwnd_test(struct tcp_sock *tp, struct sk_buff *skb)
{
	u32 in_flight, cwnd;

	/* Don't be strict about the congestion window for the final FIN.  */
	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
		return 1;

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

	return 0;
}

/* This must be invoked the first time we consider transmitting
 * SKB onto the wire.
 */
S
Stephen Hemminger 已提交
904
static int tcp_init_tso_segs(struct sock *sk, struct sk_buff *skb, unsigned int mss_now)
905 906 907
{
	int tso_segs = tcp_skb_pcount(skb);

908 909 910 911
	if (!tso_segs ||
	    (tso_segs > 1 &&
	     skb_shinfo(skb)->tso_size != mss_now)) {
		tcp_set_skb_tso_segs(sk, skb, mss_now);
912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988
		tso_segs = tcp_skb_pcount(skb);
	}
	return tso_segs;
}

static inline int tcp_minshall_check(const struct tcp_sock *tp)
{
	return after(tp->snd_sml,tp->snd_una) &&
		!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,
				  const struct sk_buff *skb, 
				  unsigned mss_now, int nonagle)
{
	return (skb->len < mss_now &&
		((nonagle&TCP_NAGLE_CORK) ||
		 (!nonagle &&
		  tp->packets_out &&
		  tcp_minshall_check(tp))));
}

/* 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;

	/* Don't use the nagle rule for urgent data (or for the final FIN).  */
	if (tp->urg_mode ||
	    (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? */
static inline int tcp_snd_wnd_test(struct tcp_sock *tp, struct sk_buff *skb, unsigned int cur_mss)
{
	u32 end_seq = TCP_SKB_CB(skb)->end_seq;

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

	return !after(end_seq, tp->snd_una + tp->snd_wnd);
}

/* This checks if the data bearing packet SKB (usually sk->sk_send_head)
 * 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;

989
	tcp_init_tso_segs(sk, skb, cur_mss);
990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025

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

	cwnd_quota = tcp_cwnd_test(tp, skb);
	if (cwnd_quota &&
	    !tcp_snd_wnd_test(tp, skb, cur_mss))
		cwnd_quota = 0;

	return cwnd_quota;
}

static inline int tcp_skb_is_last(const struct sock *sk, 
				  const struct sk_buff *skb)
{
	return skb->next == (struct sk_buff *)&sk->sk_write_queue;
}

int tcp_may_send_now(struct sock *sk, struct tcp_sock *tp)
{
	struct sk_buff *skb = sk->sk_send_head;

	return (skb &&
		tcp_snd_test(sk, skb, tcp_current_mss(sk, 1),
			     (tcp_skb_is_last(sk, skb) ?
			      TCP_NAGLE_PUSH :
			      tp->nonagle)));
}

/* 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).
 */
1026
static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, unsigned int mss_now)
1027 1028 1029 1030 1031 1032
{
	struct sk_buff *buff;
	int nlen = skb->len - len;
	u16 flags;

	/* All of a TSO frame must be composed of paged data.  */
1033 1034
	if (skb->len != skb->data_len)
		return tcp_fragment(sk, skb, len, mss_now);
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059

	buff = sk_stream_alloc_pskb(sk, 0, 0, GFP_ATOMIC);
	if (unlikely(buff == NULL))
		return -ENOMEM;

	buff->truesize = nlen;
	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;
	TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
	TCP_SKB_CB(buff)->flags = flags;

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

	buff->ip_summed = skb->ip_summed = CHECKSUM_HW;
	skb_split(skb, buff, len);

	/* Fix up tso_factor for both original and new SKB.  */
1060 1061
	tcp_set_skb_tso_segs(sk, skb, mss_now);
	tcp_set_skb_tso_segs(sk, buff, mss_now);
1062 1063 1064

	/* Link BUFF into the send queue. */
	skb_header_release(buff);
D
David S. Miller 已提交
1065
	__skb_append(skb, buff, &sk->sk_write_queue);
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076

	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.
 */
static int tcp_tso_should_defer(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb)
{
1077
	const struct inet_connection_sock *icsk = inet_csk(sk);
1078 1079 1080 1081 1082
	u32 send_win, cong_win, limit, in_flight;

	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
		return 0;

1083
	if (icsk->icsk_ca_state != TCP_CA_Open)
1084 1085
		return 0;

1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
	in_flight = tcp_packets_in_flight(tp);

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

	send_win = (tp->snd_una + tp->snd_wnd) - TCP_SKB_CB(skb)->seq;

	/* 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);

1098 1099 1100 1101
	/* If a full-sized TSO skb can be sent, do it. */
	if (limit >= 65536)
		return 0;

1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124
	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)
			return 0;
	} 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)
			return 0;
	}

	/* Ok, it looks like it is advisable to defer.  */
	return 1;
}

J
John Heffner 已提交
1125 1126 1127 1128 1129 1130 1131 1132 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 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
/* Create a new MTU probe if we are ready.
 * Returns 0 if we should wait to probe (no cwnd available),
 *         1 if a probe was sent,
 *         -1 otherwise */
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;
	unsigned int pif;
	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 ||
	    tp->rx_opt.eff_sacks)
		return -1;

	/* Very simple search strategy: just double the MSS. */
	mss_now = tcp_current_mss(sk, 0);
	probe_size = 2*tp->mss_cache;
	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? */
	len = 0;
	if ((skb = sk->sk_send_head) == NULL)
		return -1;
	while ((len += skb->len) < probe_size && !tcp_skb_is_last(sk, skb))
		skb = skb->next;
	if (len < probe_size)
		return -1;

	/* Receive window check. */
	if (after(TCP_SKB_CB(skb)->seq + probe_size, tp->snd_una + tp->snd_wnd)) {
		if (tp->snd_wnd < probe_size)
			return -1;
		else
			return 0;
	}

	/* Do we need to wait to drain cwnd? */
	pif = tcp_packets_in_flight(tp);
	if (pif + 2 > tp->snd_cwnd) {
		/* With no packets in flight, don't stall. */
		if (pif == 0)
			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;
	sk_charge_skb(sk, nskb);

	skb = sk->sk_send_head;
	__skb_insert(nskb, skb->prev, skb, &sk->sk_write_queue);
	sk->sk_send_head = nskb;

	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;
	if (skb->ip_summed == CHECKSUM_HW)
		nskb->ip_summed = CHECKSUM_HW;

	len = 0;
	while (len < probe_size) {
		next = skb->next;

		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,
			                 skb_put(nskb, copy), copy, nskb->csum);

		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;
			__skb_unlink(skb, &sk->sk_write_queue);
			sk_stream_free_skb(sk, skb);
		} else {
			TCP_SKB_CB(nskb)->flags |= TCP_SKB_CB(skb)->flags &
			                           ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
			if (!skb_shinfo(skb)->nr_frags) {
				skb_pull(skb, copy);
				if (skb->ip_summed != CHECKSUM_HW)
					skb->csum = csum_partial(skb->data, skb->len, 0);
			} else {
				__pskb_trim_head(skb, copy);
				tcp_set_skb_tso_segs(sk, skb, mss_now);
			}
			TCP_SKB_CB(skb)->seq += copy;
		}

		len += copy;
		skb = next;
	}
	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
		* effectively two packets. */
		tp->snd_cwnd--;
		update_send_head(sk, tp, nskb);

		icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len);
1249 1250
		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
1251 1252 1253 1254 1255 1256 1257 1258

		return 1;
	}

	return -1;
}


L
Linus Torvalds 已提交
1259 1260 1261 1262 1263 1264 1265
/* This routine writes packets to the network.  It advances the
 * send_head.  This happens as incoming acks open up the remote
 * window for us.
 *
 * Returns 1, if no segments are in flight and we have queued segments, but
 * cannot send anything now because of SWS or another problem.
 */
1266
static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
L
Linus Torvalds 已提交
1267 1268
{
	struct tcp_sock *tp = tcp_sk(sk);
1269
	struct sk_buff *skb;
1270 1271
	unsigned int tso_segs, sent_pkts;
	int cwnd_quota;
J
John Heffner 已提交
1272
	int result;
L
Linus Torvalds 已提交
1273 1274 1275 1276 1277

	/* 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.
	 */
1278 1279
	if (unlikely(sk->sk_state == TCP_CLOSE))
		return 0;
L
Linus Torvalds 已提交
1280

1281
	sent_pkts = 0;
J
John Heffner 已提交
1282 1283 1284 1285 1286 1287 1288 1289

	/* Do MTU probing. */
	if ((result = tcp_mtu_probe(sk)) == 0) {
		return 0;
	} else if (result > 0) {
		sent_pkts = 1;
	}

1290
	while ((skb = sk->sk_send_head)) {
1291 1292
		unsigned int limit;

1293
		tso_segs = tcp_init_tso_segs(sk, skb, mss_now);
1294
		BUG_ON(!tso_segs);
1295

1296 1297 1298 1299 1300 1301 1302
		cwnd_quota = tcp_cwnd_test(tp, skb);
		if (!cwnd_quota)
			break;

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

1303 1304 1305 1306 1307 1308 1309 1310 1311
		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 {
			if (tcp_tso_should_defer(sk, tp, skb))
				break;
		}
1312

1313
		limit = mss_now;
1314
		if (tso_segs > 1) {
1315 1316
			limit = tcp_window_allows(tp, skb,
						  mss_now, cwnd_quota);
1317 1318 1319

			if (skb->len < limit) {
				unsigned int trim = skb->len % mss_now;
1320

1321 1322 1323
				if (trim)
					limit = skb->len - trim;
			}
1324
		}
L
Linus Torvalds 已提交
1325

1326 1327 1328 1329
		if (skb->len > limit &&
		    unlikely(tso_fragment(sk, skb, limit, mss_now)))
			break;

1330
		TCP_SKB_CB(skb)->when = tcp_time_stamp;
1331

1332
		if (unlikely(tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC)))
1333
			break;
L
Linus Torvalds 已提交
1334

1335 1336 1337 1338
		/* Advance the send_head.  This one is sent out.
		 * This call will increment packets_out.
		 */
		update_send_head(sk, tp, skb);
L
Linus Torvalds 已提交
1339

1340
		tcp_minshall_update(tp, mss_now, skb);
1341
		sent_pkts++;
1342
	}
L
Linus Torvalds 已提交
1343

1344
	if (likely(sent_pkts)) {
1345 1346
		tcp_cwnd_validate(sk, tp);
		return 0;
L
Linus Torvalds 已提交
1347
	}
1348
	return !tp->packets_out && sk->sk_send_head;
L
Linus Torvalds 已提交
1349 1350
}

1351 1352 1353 1354 1355
/* 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.
 */
void __tcp_push_pending_frames(struct sock *sk, struct tcp_sock *tp,
1356
			       unsigned int cur_mss, int nonagle)
1357 1358 1359 1360
{
	struct sk_buff *skb = sk->sk_send_head;

	if (skb) {
1361
		if (tcp_write_xmit(sk, cur_mss, nonagle))
1362 1363 1364 1365
			tcp_check_probe_timer(sk, tp);
	}
}

1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
/* 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)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb = sk->sk_send_head;
	unsigned int tso_segs, cwnd_quota;

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

1377
	tso_segs = tcp_init_tso_segs(sk, skb, mss_now);
1378 1379 1380
	cwnd_quota = tcp_snd_test(sk, skb, mss_now, TCP_NAGLE_PUSH);

	if (likely(cwnd_quota)) {
1381 1382
		unsigned int limit;

1383 1384
		BUG_ON(!tso_segs);

1385
		limit = mss_now;
1386
		if (tso_segs > 1) {
1387 1388
			limit = tcp_window_allows(tp, skb,
						  mss_now, cwnd_quota);
1389 1390 1391 1392 1393 1394 1395 1396 1397

			if (skb->len < limit) {
				unsigned int trim = skb->len % mss_now;

				if (trim)
					limit = skb->len - trim;
			}
		}

1398 1399 1400 1401
		if (skb->len > limit &&
		    unlikely(tso_fragment(sk, skb, limit, mss_now)))
			return;

1402 1403 1404
		/* Send it out now. */
		TCP_SKB_CB(skb)->when = tcp_time_stamp;

1405
		if (likely(!tcp_transmit_skb(sk, skb, 1, sk->sk_allocation))) {
1406 1407 1408 1409 1410 1411 1412
			update_send_head(sk, tp, skb);
			tcp_cwnd_validate(sk, tp);
			return;
		}
	}
}

L
Linus Torvalds 已提交
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466
/* This function returns the amount that we can raise the
 * usable window based on the following constraints
 *  
 * 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.
 * 
 * Of course, if the sender implements sender side SWS prevention
 * then this will not be a problem.
 * 
 * BSD seems to make the following compromise:
 * 
 *	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)
{
1467
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1468
	struct tcp_sock *tp = tcp_sk(sk);
S
Stephen Hemminger 已提交
1469
	/* MSS for the peer's data.  Previous versions used mss_clamp
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474
	 * 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
	 */
1475
	int mss = icsk->icsk_ack.rcv_mss;
L
Linus Torvalds 已提交
1476 1477 1478 1479 1480 1481 1482 1483
	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)
		mss = full_space; 

	if (free_space < full_space/2) {
1484
		icsk->icsk_ack.quick = 0;
L
Linus Torvalds 已提交
1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557

		if (tcp_memory_pressure)
			tp->rcv_ssthresh = min(tp->rcv_ssthresh, 4U*tp->advmss);

		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)
			window = (free_space/mss)*mss;
	}

	return window;
}

/* Attempt to collapse two adjacent SKB's during retransmission. */
static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int mss_now)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *next_skb = skb->next;

	/* The first test we must make is that neither of these two
	 * SKB's are still referenced by someone else.
	 */
	if (!skb_cloned(skb) && !skb_cloned(next_skb)) {
		int skb_size = skb->len, next_skb_size = next_skb->len;
		u16 flags = TCP_SKB_CB(skb)->flags;

		/* Also punt if next skb has been SACK'd. */
		if(TCP_SKB_CB(next_skb)->sacked & TCPCB_SACKED_ACKED)
			return;

		/* Next skb is out of window. */
		if (after(TCP_SKB_CB(next_skb)->end_seq, tp->snd_una+tp->snd_wnd))
			return;

		/* Punt if not enough space exists in the first SKB for
		 * the data in the second, or the total combined payload
		 * would exceed the MSS.
		 */
		if ((next_skb_size > skb_tailroom(skb)) ||
		    ((skb_size + next_skb_size) > mss_now))
			return;

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

1558 1559 1560 1561
		/* changing transmit queue under us so clear hints */
		clear_all_retrans_hints(tp);

		/* Ok.	We will be able to collapse the packet. */
D
David S. Miller 已提交
1562
		__skb_unlink(next_skb, &sk->sk_write_queue);
L
Linus Torvalds 已提交
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609

		memcpy(skb_put(skb, next_skb_size), next_skb->data, next_skb_size);

		if (next_skb->ip_summed == CHECKSUM_HW)
			skb->ip_summed = CHECKSUM_HW;

		if (skb->ip_summed != CHECKSUM_HW)
			skb->csum = csum_block_add(skb->csum, next_skb->csum, skb_size);

		/* Update sequence range on original skb. */
		TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;

		/* Merge over control information. */
		flags |= TCP_SKB_CB(next_skb)->flags; /* This moves PSH/FIN etc. over */
		TCP_SKB_CB(skb)->flags = flags;

		/* 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|TCPCB_AT_TAIL);
		if (TCP_SKB_CB(next_skb)->sacked&TCPCB_SACKED_RETRANS)
			tp->retrans_out -= tcp_skb_pcount(next_skb);
		if (TCP_SKB_CB(next_skb)->sacked&TCPCB_LOST) {
			tp->lost_out -= tcp_skb_pcount(next_skb);
			tp->left_out -= tcp_skb_pcount(next_skb);
		}
		/* Reno case is special. Sigh... */
		if (!tp->rx_opt.sack_ok && tp->sacked_out) {
			tcp_dec_pcount_approx(&tp->sacked_out, next_skb);
			tp->left_out -= tcp_skb_pcount(next_skb);
		}

		/* Not quite right: it can be > snd.fack, but
		 * it is better to underestimate fackets.
		 */
		tcp_dec_pcount_approx(&tp->fackets_out, next_skb);
		tcp_packets_out_dec(tp, next_skb);
		sk_stream_free_skb(sk, next_skb);
	}
}

/* Do a simple retransmit without using the backoff mechanisms in
 * tcp_timer. This is used for path mtu discovery. 
 * The socket is already locked here.
 */ 
void tcp_simple_retransmit(struct sock *sk)
{
1610
	const struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;
	unsigned int mss = tcp_current_mss(sk, 0);
	int lost = 0;

	sk_stream_for_retrans_queue(skb, sk) {
		if (skb->len > mss && 
		    !(TCP_SKB_CB(skb)->sacked&TCPCB_SACKED_ACKED)) {
			if (TCP_SKB_CB(skb)->sacked&TCPCB_SACKED_RETRANS) {
				TCP_SKB_CB(skb)->sacked &= ~TCPCB_SACKED_RETRANS;
				tp->retrans_out -= tcp_skb_pcount(skb);
			}
			if (!(TCP_SKB_CB(skb)->sacked&TCPCB_LOST)) {
				TCP_SKB_CB(skb)->sacked |= TCPCB_LOST;
				tp->lost_out += tcp_skb_pcount(skb);
				lost = 1;
			}
		}
	}

1631 1632
	clear_all_retrans_hints(tp);

L
Linus Torvalds 已提交
1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
	if (!lost)
		return;

	tcp_sync_left_out(tp);

 	/* Don't muck with the congestion window here.
	 * Reason is that we do not increase amount of _data_
	 * in network, but units changed and effective
	 * cwnd/ssthresh really reduced now.
	 */
1643
	if (icsk->icsk_ca_state != TCP_CA_Loss) {
L
Linus Torvalds 已提交
1644
		tp->high_seq = tp->snd_nxt;
1645
		tp->snd_ssthresh = tcp_current_ssthresh(sk);
L
Linus Torvalds 已提交
1646 1647
		tp->prior_ssthresh = 0;
		tp->undo_marker = 0;
1648
		tcp_set_ca_state(sk, TCP_CA_Loss);
L
Linus Torvalds 已提交
1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
	}
	tcp_xmit_retransmit_queue(sk);
}

/* 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
1660
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1661 1662 1663
 	unsigned int cur_mss = tcp_current_mss(sk, 0);
	int err;

J
John Heffner 已提交
1664 1665 1666 1667 1668
	/* Inconslusive MTU probe */
	if (icsk->icsk_mtup.probe_size) {
		icsk->icsk_mtup.probe_size = 0;
	}

L
Linus Torvalds 已提交
1669
	/* Do not sent more than we queued. 1/4 is reserved for possible
S
Stephen Hemminger 已提交
1670
	 * copying overhead: fragmentation, tunneling, mangling etc.
L
Linus Torvalds 已提交
1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692
	 */
	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;
	}

	/* 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.
	 */
	if (!before(TCP_SKB_CB(skb)->seq, tp->snd_una+tp->snd_wnd)
	    && TCP_SKB_CB(skb)->seq != tp->snd_una)
		return -EAGAIN;

	if (skb->len > cur_mss) {
1693
		if (tcp_fragment(sk, skb, cur_mss, cur_mss))
L
Linus Torvalds 已提交
1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706
			return -ENOMEM; /* We'll try again later. */
	}

	/* Collapse two adjacent packets if worthwhile and we can. */
	if(!(TCP_SKB_CB(skb)->flags & TCPCB_FLAG_SYN) &&
	   (skb->len < (cur_mss >> 1)) &&
	   (skb->next != sk->sk_send_head) &&
	   (skb->next != (struct sk_buff *)&sk->sk_write_queue) &&
	   (skb_shinfo(skb)->nr_frags == 0 && skb_shinfo(skb->next)->nr_frags == 0) &&
	   (tcp_skb_pcount(skb) == 1 && tcp_skb_pcount(skb->next) == 1) &&
	   (sysctl_tcp_retrans_collapse != 0))
		tcp_retrans_try_collapse(sk, skb, cur_mss);

1707
	if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk))
L
Linus Torvalds 已提交
1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730
		return -EHOSTUNREACH; /* Routing failure or similar. */

	/* 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.
	 */
	if(skb->len > 0 &&
	   (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN) &&
	   tp->snd_una == (TCP_SKB_CB(skb)->end_seq - 1)) {
		if (!pskb_trim(skb, 0)) {
			TCP_SKB_CB(skb)->seq = TCP_SKB_CB(skb)->end_seq - 1;
			skb_shinfo(skb)->tso_segs = 1;
			skb_shinfo(skb)->tso_size = 0;
			skb->ip_summed = CHECKSUM_NONE;
			skb->csum = 0;
		}
	}

	/* 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;

1731
	err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
L
Linus Torvalds 已提交
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771

	if (err == 0) {
		/* Update global TCP statistics. */
		TCP_INC_STATS(TCP_MIB_RETRANSSEGS);

		tp->total_retrans++;

#if FASTRETRANS_DEBUG > 0
		if (TCP_SKB_CB(skb)->sacked&TCPCB_SACKED_RETRANS) {
			if (net_ratelimit())
				printk(KERN_DEBUG "retrans_out leaked.\n");
		}
#endif
		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;
}

/* 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)
{
1772
	const struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1773 1774
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;
1775 1776 1777 1778 1779 1780 1781 1782 1783
	int packet_cnt;

	if (tp->retransmit_skb_hint) {
		skb = tp->retransmit_skb_hint;
		packet_cnt = tp->retransmit_cnt_hint;
	}else{
		skb = sk->sk_write_queue.next;
		packet_cnt = 0;
	}
L
Linus Torvalds 已提交
1784 1785

	/* First pass: retransmit lost packets. */
1786 1787
	if (tp->lost_out) {
		sk_stream_for_retrans_queue_from(skb, sk) {
L
Linus Torvalds 已提交
1788 1789
			__u8 sacked = TCP_SKB_CB(skb)->sacked;

1790 1791 1792 1793
			/* we could do better than to assign each time */
			tp->retransmit_skb_hint = skb;
			tp->retransmit_cnt_hint = packet_cnt;

L
Linus Torvalds 已提交
1794 1795 1796 1797 1798 1799 1800 1801 1802 1803
			/* 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;

1804
			if (sacked & TCPCB_LOST) {
L
Linus Torvalds 已提交
1805
				if (!(sacked&(TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))) {
1806 1807
					if (tcp_retransmit_skb(sk, skb)) {
						tp->retransmit_skb_hint = NULL;
L
Linus Torvalds 已提交
1808
						return;
1809
					}
1810
					if (icsk->icsk_ca_state != TCP_CA_Loss)
L
Linus Torvalds 已提交
1811 1812 1813 1814 1815 1816
						NET_INC_STATS_BH(LINUX_MIB_TCPFASTRETRANS);
					else
						NET_INC_STATS_BH(LINUX_MIB_TCPSLOWSTARTRETRANS);

					if (skb ==
					    skb_peek(&sk->sk_write_queue))
1817
						inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
1818 1819
									  inet_csk(sk)->icsk_rto,
									  TCP_RTO_MAX);
L
Linus Torvalds 已提交
1820 1821
				}

1822 1823
				packet_cnt += tcp_skb_pcount(skb);
				if (packet_cnt >= tp->lost_out)
L
Linus Torvalds 已提交
1824 1825 1826 1827 1828 1829 1830 1831
					break;
			}
		}
	}

	/* OK, demanded retransmission is finished. */

	/* Forward retransmissions are possible only during Recovery. */
1832
	if (icsk->icsk_ca_state != TCP_CA_Recovery)
L
Linus Torvalds 已提交
1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848
		return;

	/* No forward retransmissions in Reno are possible. */
	if (!tp->rx_opt.sack_ok)
		return;

	/* 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.
	 */

	if (tcp_may_send_now(sk, tp))
		return;

1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859
	if (tp->forward_skb_hint) {
		skb = tp->forward_skb_hint;
		packet_cnt = tp->forward_cnt_hint;
	} else{
		skb = sk->sk_write_queue.next;
		packet_cnt = 0;
	}

	sk_stream_for_retrans_queue_from(skb, sk) {
		tp->forward_cnt_hint = packet_cnt;
		tp->forward_skb_hint = skb;
L
Linus Torvalds 已提交
1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876

		/* Similar to the retransmit loop above we
		 * can pretend that the retransmitted SKB
		 * we send out here will be composed of one
		 * real MSS sized packet because tcp_retransmit_skb()
		 * will fragment it if necessary.
		 */
		if (++packet_cnt > tp->fackets_out)
			break;

		if (tcp_packets_in_flight(tp) >= tp->snd_cwnd)
			break;

		if (TCP_SKB_CB(skb)->sacked & TCPCB_TAGBITS)
			continue;

		/* Ok, retransmit it. */
1877 1878
		if (tcp_retransmit_skb(sk, skb)) {
			tp->forward_skb_hint = NULL;
L
Linus Torvalds 已提交
1879
			break;
1880
		}
L
Linus Torvalds 已提交
1881 1882

		if (skb == skb_peek(&sk->sk_write_queue))
1883 1884 1885
			inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
						  inet_csk(sk)->icsk_rto,
						  TCP_RTO_MAX);
L
Linus Torvalds 已提交
1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913

		NET_INC_STATS_BH(LINUX_MIB_TCPFORWARDRETRANS);
	}
}


/* 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)
{
	struct tcp_sock *tp = tcp_sk(sk);	
	struct sk_buff *skb = skb_peek_tail(&sk->sk_write_queue);
	int mss_now;
	
	/* Optimization, tack on the FIN if we have a queue of
	 * unsent frames.  But be careful about outgoing SACKS
	 * and IP options.
	 */
	mss_now = tcp_current_mss(sk, 1);

	if (sk->sk_send_head != NULL) {
		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 (;;) {
1914
			skb = alloc_skb_fclone(MAX_TCP_HEADER, GFP_KERNEL);
L
Linus Torvalds 已提交
1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940
			if (skb)
				break;
			yield();
		}

		/* Reserve space for headers and prepare control bits. */
		skb_reserve(skb, MAX_TCP_HEADER);
		skb->csum = 0;
		TCP_SKB_CB(skb)->flags = (TCPCB_FLAG_ACK | TCPCB_FLAG_FIN);
		TCP_SKB_CB(skb)->sacked = 0;
		skb_shinfo(skb)->tso_segs = 1;
		skb_shinfo(skb)->tso_size = 0;

		/* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */
		TCP_SKB_CB(skb)->seq = tp->write_seq;
		TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq + 1;
		tcp_queue_skb(sk, skb);
	}
	__tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_OFF);
}

/* 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
 * by draft-ietf-tcpimpl-prob-03.txt section 3.10.  -DaveM
 */
A
Al Viro 已提交
1941
void tcp_send_active_reset(struct sock *sk, gfp_t priority)
L
Linus Torvalds 已提交
1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *skb;

	/* NOTE: No TCP options attached and we never retransmit this. */
	skb = alloc_skb(MAX_TCP_HEADER, priority);
	if (!skb) {
		NET_INC_STATS(LINUX_MIB_TCPABORTFAILED);
		return;
	}

	/* Reserve space for headers and prepare control bits. */
	skb_reserve(skb, MAX_TCP_HEADER);
	skb->csum = 0;
	TCP_SKB_CB(skb)->flags = (TCPCB_FLAG_ACK | TCPCB_FLAG_RST);
	TCP_SKB_CB(skb)->sacked = 0;
	skb_shinfo(skb)->tso_segs = 1;
	skb_shinfo(skb)->tso_size = 0;

	/* Send it off. */
	TCP_SKB_CB(skb)->seq = tcp_acceptable_seq(sk, tp);
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq;
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
1965
	if (tcp_transmit_skb(sk, skb, 0, priority))
L
Linus Torvalds 已提交
1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
		NET_INC_STATS(LINUX_MIB_TCPABORTFAILED);
}

/* WARNING: This routine must only be called when we have already sent
 * 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)
{
	struct sk_buff* skb;

	skb = skb_peek(&sk->sk_write_queue);
	if (skb == NULL || !(TCP_SKB_CB(skb)->flags&TCPCB_FLAG_SYN)) {
		printk(KERN_DEBUG "tcp_send_synack: wrong queue state\n");
		return -EFAULT;
	}
	if (!(TCP_SKB_CB(skb)->flags&TCPCB_FLAG_ACK)) {
		if (skb_cloned(skb)) {
			struct sk_buff *nskb = skb_copy(skb, GFP_ATOMIC);
			if (nskb == NULL)
				return -ENOMEM;
			__skb_unlink(skb, &sk->sk_write_queue);
			skb_header_release(nskb);
			__skb_queue_head(&sk->sk_write_queue, nskb);
			sk_stream_free_skb(sk, skb);
			sk_charge_skb(sk, nskb);
			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;
2000
	return tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
L
Linus Torvalds 已提交
2001 2002 2003 2004 2005 2006
}

/*
 * Prepare a SYN-ACK.
 */
struct sk_buff * tcp_make_synack(struct sock *sk, struct dst_entry *dst,
2007
				 struct request_sock *req)
L
Linus Torvalds 已提交
2008
{
2009
	struct inet_request_sock *ireq = inet_rsk(req);
L
Linus Torvalds 已提交
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
	struct tcp_sock *tp = tcp_sk(sk);
	struct tcphdr *th;
	int tcp_header_size;
	struct sk_buff *skb;

	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);

	skb->dst = dst_clone(dst);

	tcp_header_size = (sizeof(struct tcphdr) + TCPOLEN_MSS +
2025 2026
			   (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0) +
			   (ireq->wscale_ok ? TCPOLEN_WSCALE_ALIGNED : 0) +
L
Linus Torvalds 已提交
2027
			   /* SACK_PERM is in the place of NOP NOP of TS */
2028
			   ((ireq->sack_ok && !ireq->tstamp_ok) ? TCPOLEN_SACKPERM_ALIGNED : 0));
L
Linus Torvalds 已提交
2029 2030 2031 2032 2033 2034
	skb->h.th = th = (struct tcphdr *) skb_push(skb, tcp_header_size);

	memset(th, 0, sizeof(struct tcphdr));
	th->syn = 1;
	th->ack = 1;
	if (dst->dev->features&NETIF_F_TSO)
2035
		ireq->ecn_ok = 0;
L
Linus Torvalds 已提交
2036 2037
	TCP_ECN_make_synack(req, th);
	th->source = inet_sk(sk)->sport;
2038 2039
	th->dest = ireq->rmt_port;
	TCP_SKB_CB(skb)->seq = tcp_rsk(req)->snt_isn;
L
Linus Torvalds 已提交
2040 2041 2042 2043 2044
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq + 1;
	TCP_SKB_CB(skb)->sacked = 0;
	skb_shinfo(skb)->tso_segs = 1;
	skb_shinfo(skb)->tso_size = 0;
	th->seq = htonl(TCP_SKB_CB(skb)->seq);
2045
	th->ack_seq = htonl(tcp_rsk(req)->rcv_isn + 1);
L
Linus Torvalds 已提交
2046 2047 2048 2049 2050 2051
	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), 
2052
			dst_metric(dst, RTAX_ADVMSS) - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0),
L
Linus Torvalds 已提交
2053 2054
			&req->rcv_wnd,
			&req->window_clamp,
2055
			ireq->wscale_ok,
L
Linus Torvalds 已提交
2056
			&rcv_wscale);
2057
		ireq->rcv_wscale = rcv_wscale; 
L
Linus Torvalds 已提交
2058 2059 2060 2061 2062 2063
	}

	/* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */
	th->window = htons(req->rcv_wnd);

	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2064 2065
	tcp_syn_build_options((__u32 *)(th + 1), dst_metric(dst, RTAX_ADVMSS), ireq->tstamp_ok,
			      ireq->sack_ok, ireq->wscale_ok, ireq->rcv_wscale,
L
Linus Torvalds 已提交
2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077
			      TCP_SKB_CB(skb)->when,
			      req->ts_recent);

	skb->csum = 0;
	th->doff = (tcp_header_size >> 2);
	TCP_INC_STATS(TCP_MIB_OUTSEGS);
	return skb;
}

/* 
 * Do all connect socket setups that can be done AF independent.
 */ 
S
Stephen Hemminger 已提交
2078
static void tcp_connect_init(struct sock *sk)
L
Linus Torvalds 已提交
2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093
{
	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) +
		(sysctl_tcp_timestamps ? TCPOLEN_TSTAMP_ALIGNED : 0);

	/* 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
2094
	tcp_mtup_init(sk);
L
Linus Torvalds 已提交
2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121
	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);
	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,
				  sysctl_tcp_window_scaling,
				  &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;
	tcp_init_wl(tp, tp->write_seq, 0);
	tp->snd_una = tp->write_seq;
	tp->snd_sml = tp->write_seq;
	tp->rcv_nxt = 0;
	tp->rcv_wup = 0;
	tp->copied_seq = 0;

2122 2123
	inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT;
	inet_csk(sk)->icsk_retransmits = 0;
L
Linus Torvalds 已提交
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136
	tcp_clear_retrans(tp);
}

/*
 * Build a SYN and send it off.
 */ 
int tcp_connect(struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	struct sk_buff *buff;

	tcp_connect_init(sk);

2137
	buff = alloc_skb_fclone(MAX_TCP_HEADER + 15, sk->sk_allocation);
L
Linus Torvalds 已提交
2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161
	if (unlikely(buff == NULL))
		return -ENOBUFS;

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

	TCP_SKB_CB(buff)->flags = TCPCB_FLAG_SYN;
	TCP_ECN_send_syn(sk, tp, buff);
	TCP_SKB_CB(buff)->sacked = 0;
	skb_shinfo(buff)->tso_segs = 1;
	skb_shinfo(buff)->tso_size = 0;
	buff->csum = 0;
	TCP_SKB_CB(buff)->seq = tp->write_seq++;
	TCP_SKB_CB(buff)->end_seq = tp->write_seq;
	tp->snd_nxt = tp->write_seq;
	tp->pushed_seq = tp->write_seq;

	/* Send it off. */
	TCP_SKB_CB(buff)->when = tcp_time_stamp;
	tp->retrans_stamp = TCP_SKB_CB(buff)->when;
	skb_header_release(buff);
	__skb_queue_tail(&sk->sk_write_queue, buff);
	sk_charge_skb(sk, buff);
	tp->packets_out += tcp_skb_pcount(buff);
2162
	tcp_transmit_skb(sk, buff, 1, GFP_KERNEL);
L
Linus Torvalds 已提交
2163 2164 2165
	TCP_INC_STATS(TCP_MIB_ACTIVEOPENS);

	/* Timer for repeating the SYN until an answer. */
2166 2167
	inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
				  inet_csk(sk)->icsk_rto, TCP_RTO_MAX);
L
Linus Torvalds 已提交
2168 2169 2170 2171 2172 2173 2174 2175 2176
	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)
{
2177 2178
	struct inet_connection_sock *icsk = inet_csk(sk);
	int ato = icsk->icsk_ack.ato;
L
Linus Torvalds 已提交
2179 2180 2181
	unsigned long timeout;

	if (ato > TCP_DELACK_MIN) {
2182
		const struct tcp_sock *tp = tcp_sk(sk);
L
Linus Torvalds 已提交
2183 2184
		int max_ato = HZ/2;

2185
		if (icsk->icsk_ack.pingpong || (icsk->icsk_ack.pending & ICSK_ACK_PUSHED))
L
Linus Torvalds 已提交
2186 2187 2188 2189 2190
			max_ato = TCP_DELACK_MAX;

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

		/* If some rtt estimate is known, use it to bound delayed ack.
2191
		 * Do not use inet_csk(sk)->icsk_rto here, use results of rtt measurements
L
Linus Torvalds 已提交
2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207
		 * directly.
		 */
		if (tp->srtt) {
			int rtt = max(tp->srtt>>3, TCP_DELACK_MIN);

			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. */
2208
	if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) {
L
Linus Torvalds 已提交
2209 2210 2211
		/* If delack timer was blocked or is about to expire,
		 * send ACK now.
		 */
2212 2213
		if (icsk->icsk_ack.blocked ||
		    time_before_eq(icsk->icsk_ack.timeout, jiffies + (ato >> 2))) {
L
Linus Torvalds 已提交
2214 2215 2216 2217
			tcp_send_ack(sk);
			return;
		}

2218 2219
		if (!time_before(timeout, icsk->icsk_ack.timeout))
			timeout = icsk->icsk_ack.timeout;
L
Linus Torvalds 已提交
2220
	}
2221 2222 2223
	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
2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239
}

/* This routine sends an ack and also updates the window. */
void tcp_send_ack(struct sock *sk)
{
	/* If we have been reset, we may not send again. */
	if (sk->sk_state != TCP_CLOSE) {
		struct tcp_sock *tp = tcp_sk(sk);
		struct sk_buff *buff;

		/* 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) {
2240 2241
			inet_csk_schedule_ack(sk);
			inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN;
2242 2243
			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
						  TCP_DELACK_MAX, TCP_RTO_MAX);
L
Linus Torvalds 已提交
2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257
			return;
		}

		/* Reserve space for headers and prepare control bits. */
		skb_reserve(buff, MAX_TCP_HEADER);
		buff->csum = 0;
		TCP_SKB_CB(buff)->flags = TCPCB_FLAG_ACK;
		TCP_SKB_CB(buff)->sacked = 0;
		skb_shinfo(buff)->tso_segs = 1;
		skb_shinfo(buff)->tso_size = 0;

		/* Send it off, this clears delayed acks for us. */
		TCP_SKB_CB(buff)->seq = TCP_SKB_CB(buff)->end_seq = tcp_acceptable_seq(sk, tp);
		TCP_SKB_CB(buff)->when = tcp_time_stamp;
2258
		tcp_transmit_skb(sk, buff, 0, GFP_ATOMIC);
L
Linus Torvalds 已提交
2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297
	}
}

/* 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);
	if (skb == NULL) 
		return -1;

	/* Reserve space for headers and set control bits. */
	skb_reserve(skb, MAX_TCP_HEADER);
	skb->csum = 0;
	TCP_SKB_CB(skb)->flags = TCPCB_FLAG_ACK;
	TCP_SKB_CB(skb)->sacked = urgent;
	skb_shinfo(skb)->tso_segs = 1;
	skb_shinfo(skb)->tso_size = 0;

	/* Use a previous sequence.  This should cause the other
	 * end to send an ack.  Don't queue or clone SKB, just
	 * send it.
	 */
	TCP_SKB_CB(skb)->seq = urgent ? tp->snd_una : tp->snd_una - 1;
	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq;
	TCP_SKB_CB(skb)->when = tcp_time_stamp;
2298
	return tcp_transmit_skb(sk, skb, 0, GFP_ATOMIC);
L
Linus Torvalds 已提交
2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323
}

int tcp_write_wakeup(struct sock *sk)
{
	if (sk->sk_state != TCP_CLOSE) {
		struct tcp_sock *tp = tcp_sk(sk);
		struct sk_buff *skb;

		if ((skb = sk->sk_send_head) != NULL &&
		    before(TCP_SKB_CB(skb)->seq, tp->snd_una+tp->snd_wnd)) {
			int err;
			unsigned int mss = tcp_current_mss(sk, 0);
			unsigned int seg_size = tp->snd_una+tp->snd_wnd-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);
				TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
2324
				if (tcp_fragment(sk, skb, seg_size, mss))
L
Linus Torvalds 已提交
2325 2326
					return -1;
			} else if (!tcp_skb_pcount(skb))
2327
				tcp_set_skb_tso_segs(sk, skb, mss);
L
Linus Torvalds 已提交
2328 2329 2330

			TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
			TCP_SKB_CB(skb)->when = tcp_time_stamp;
2331
			err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
L
Linus Torvalds 已提交
2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350
			if (!err) {
				update_send_head(sk, tp, skb);
			}
			return err;
		} else {
			if (tp->urg_mode &&
			    between(tp->snd_up, tp->snd_una+1, tp->snd_una+0xFFFF))
				tcp_xmit_probe_skb(sk, TCPCB_URG);
			return tcp_xmit_probe_skb(sk, 0);
		}
	}
	return -1;
}

/* 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)
{
2351
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
2352 2353 2354 2355 2356 2357 2358
	struct tcp_sock *tp = tcp_sk(sk);
	int err;

	err = tcp_write_wakeup(sk);

	if (tp->packets_out || !sk->sk_send_head) {
		/* Cancel probe timer, if it is not required. */
2359
		icsk->icsk_probes_out = 0;
2360
		icsk->icsk_backoff = 0;
L
Linus Torvalds 已提交
2361 2362 2363 2364
		return;
	}

	if (err <= 0) {
2365 2366
		if (icsk->icsk_backoff < sysctl_tcp_retries2)
			icsk->icsk_backoff++;
2367
		icsk->icsk_probes_out++;
2368
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, 
2369 2370
					  min(icsk->icsk_rto << icsk->icsk_backoff, TCP_RTO_MAX),
					  TCP_RTO_MAX);
L
Linus Torvalds 已提交
2371 2372
	} else {
		/* If packet was not sent due to local congestion,
2373
		 * do not backoff and do not remember icsk_probes_out.
L
Linus Torvalds 已提交
2374 2375 2376 2377
		 * Let local senders to fight for local resources.
		 *
		 * Use accumulated backoff yet.
		 */
2378 2379
		if (!icsk->icsk_probes_out)
			icsk->icsk_probes_out = 1;
2380 2381
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, 
					  min(icsk->icsk_rto << icsk->icsk_backoff,
2382 2383
					      TCP_RESOURCE_PROBE_INTERVAL),
					  TCP_RTO_MAX);
L
Linus Torvalds 已提交
2384 2385 2386 2387 2388 2389 2390
	}
}

EXPORT_SYMBOL(tcp_connect);
EXPORT_SYMBOL(tcp_make_synack);
EXPORT_SYMBOL(tcp_simple_retransmit);
EXPORT_SYMBOL(tcp_sync_mss);
2391
EXPORT_SYMBOL(sysctl_tcp_tso_win_divisor);
J
John Heffner 已提交
2392
EXPORT_SYMBOL(tcp_mtup_init);