X-Git-Url: https://git.subgeniuskitty.com/unix-history/.git/blobdiff_plain/2ff61f9de862796b4b10ad48f162b5d77847f1b3..05db08ca40dcf52f15be693f25a96a593409f690:/usr/src/sys/netinet/tcp_output.c

diff --git a/usr/src/sys/netinet/tcp_output.c b/usr/src/sys/netinet/tcp_output.c
index 3199ee8110..4d150a48a6 100644
--- a/usr/src/sys/netinet/tcp_output.c
+++ b/usr/src/sys/netinet/tcp_output.c
@@ -1,200 +1,425 @@
-/*	tcp_output.c	4.17	81/11/24	*/
-
-#include "../h/param.h"
-#include "../h/systm.h"
-#include "../h/mbuf.h"
-#include "../h/socket.h"
-#include "../h/socketvar.h"
-#include "../net/inet.h"
-#include "../net/inet_pcb.h"
-#include "../net/inet_systm.h"
-#include "../net/imp.h"
-#include "../net/ip.h"
-#include "../net/ip_var.h"
-#include "../net/tcp.h"
-#include "../net/tcp_var.h"
-#include "../net/tcp_fsm.h"
-#include "/usr/include/errno.h"
-
 /*
- * Special routines to send control messages.
+ * Copyright (c) 1982, 1986 Regents of the University of California.
+ * All rights reserved.  The Berkeley software License Agreement
+ * specifies the terms and conditions for redistribution.
+ *
+ *	@(#)tcp_output.c	7.10 (Berkeley) %G%
  */
-tcp_sndctl(tp)
-	struct tcpcb *tp;
-{
-COUNT(TCP_SNDCTL);
-
-        if (tcp_send(tp))
-		return (1);
-	tcp_sndnull(tp);
-	return (0);
-}
-
-tcp_sndwin(tp)
-	struct tcpcb *tp;
-{
-	int ihave, hehas;
-COUNT(TCP_SNDWIN);
 
-	if (tp->rcv_adv) {
-		register struct socket *so = tp->t_inpcb->inp_socket;
+#include "param.h"
+#include "systm.h"
+#include "mbuf.h"
+#include "protosw.h"
+#include "socket.h"
+#include "socketvar.h"
+#include "errno.h"
 
-		ihave = so->so_rcv.sb_hiwat -
-		    (so->so_rcv.sb_cc + tp->seqcnt);
-		hehas = tp->rcv_adv - tp->rcv_nxt;
-		if ((100*(ihave-hehas)/so->so_rcv.sb_hiwat) < 35)
-			return;
-	}
-        if (tcp_send(tp))
-		return;
-	tcp_sndnull(tp);
-}
+#include "../net/route.h"
 
-tcp_sndnull(tp)
-	register struct tcpcb *tp;
-{
-COUNT(TCP_SNDNULL);
+#include "in.h"
+#include "in_pcb.h"
+#include "in_systm.h"
+#include "ip.h"
+#include "ip_var.h"
+#include "tcp.h"
+#define	TCPOUTFLAGS
+#include "tcp_fsm.h"
+#include "tcp_seq.h"
+#include "tcp_timer.h"
+#include "tcp_var.h"
+#include "tcpip.h"
+#include "tcp_debug.h"
 
-	(void) tcp_output(tp, 0, 0, (struct mbuf *)0);
-        tp->tc_flags &= ~TC_ACK_DUE;
-}
+/*
+ * Initial options.
+ */
+u_char	tcp_initopt[4] = { TCPOPT_MAXSEG, 4, 0x0, 0x0, };
 
 /*
- * Tcp segment output routine.
+ * Tcp output routine: figure out what should be sent and send it.
  */
-tcp_send(tp)
+tcp_output(tp)
 	register struct tcpcb *tp;
 {
-	register unsigned long last, wind;
 	register struct socket *so = tp->t_inpcb->inp_socket;
-	struct mbuf *m;
-	int flags = 0, forced, sent, len;
-
-COUNT(TCP_SEND);
-	tp->snd_lst = tp->snd_nxt;
-	forced = 0;
-	m = NULL;
-	if (tp->snd_nxt == tp->iss) {
-		flags |= TH_SYN;
-		tp->snd_lst++;
-	}
-	last = tp->snd_off;
-	for (m = so->so_snd.sb_mb; m != NULL; m = m->m_next)
-		last += m->m_len;
-	if (tp->snd_nxt > last) {
-		if ((tp->tc_flags&TC_SND_FIN) &&
-		    (tp->seq_fin == tp->iss || tp->snd_nxt <= tp->seq_fin)) {
-
-			flags |= TH_FIN;
-			tp->seq_fin = tp->snd_lst++;
-		}
-	} else {
-		if (tp->tc_flags&TC_SYN_ACKED) {
-			wind = tp->snd_una + tp->snd_wnd;
-			tp->snd_lst = MIN(last, wind);
-			if ((len = tp->snd_lst - tp->snd_nxt) > 1024)
-				tp->snd_lst -= len - 1024;
-			if (tp->snd_lst >= wind)
-				tp->t_persist = T_PERS;
+	register int len, win;
+	struct mbuf *m0;
+	int off, flags, error;
+	register struct mbuf *m;
+	register struct tcpiphdr *ti;
+	u_char *opt;
+	unsigned optlen = 0;
+	int idle, sendalot;
+
+	/*
+	 * Determine length of data that should be transmitted,
+	 * and flags that will be used.
+	 * If there is some data or critical controls (SYN, RST)
+	 * to send, then transmit; otherwise, investigate further.
+	 */
+	idle = (tp->snd_max == tp->snd_una);
+again:
+	sendalot = 0;
+	off = tp->snd_nxt - tp->snd_una;
+	win = MIN(tp->snd_wnd, tp->snd_cwnd);
+
+	/*
+	 * If in persist timeout with window of 0, send 1 byte.
+	 * Otherwise, if window is small but nonzero
+	 * and timer expired, we will send what we can
+	 * and go to transmit state.
+	 */
+	if (tp->t_force) {
+		if (win == 0)
+			win = 1;
+		else {
+			tp->t_timer[TCPT_PERSIST] = 0;
+			tp->t_rxtshift = 0;
 		}
-		if ((tp->tc_flags&TC_FORCE_ONE) && (tp->snd_lst == wind)) {
-			tp->snd_lst = tp->snd_nxt + 1;
-			forced = 1;
-		} else if (tp->snd_nxt >= tp->snd_lst && (tp->tc_flags&TC_SND_FIN) == 0)
+	}
+
+	len = MIN(so->so_snd.sb_cc, win) - off;
+	flags = tcp_outflags[tp->t_state];
+
+	if (len < 0) {
+		/*
+		 * If FIN has been sent but not acked,
+		 * but we haven't been called to retransmit,
+		 * len will be -1; transmit if acking, otherwise no need.
+		 * Otherwise, window shrank after we sent into it.
+		 * If window shrank to 0, cancel pending retransmit
+		 * and pull snd_nxt back to (closed) window.
+		 * We will enter persist state below.
+		 * If the window didn't close completely,
+		 * just wait for an ACK.
+		 */
+		if (flags & TH_FIN) {
+			if (tp->t_flags & TF_ACKNOW)
+				len = 0;
+			else
+				return (0);
+		} else if (win == 0) {
+			tp->t_timer[TCPT_REXMT] = 0;
+			tp->snd_nxt = tp->snd_una;
+			len = 0;
+		} else
 			return (0);
-		m = m_copy(so->so_snd.sb_mb,
-		      (int)(MAX(tp->iss+1,tp->snd_nxt) - tp->snd_off),
-		      (int)(tp->snd_lst - tp->snd_off));
-		if (tp->snd_end > tp->iss && tp->snd_end <= tp->snd_lst)
-			flags |= TH_EOL;
-		if ((tp->tc_flags&TC_SND_FIN) && !forced &&
-		    tp->snd_lst == last &&
-		    (tp->seq_fin == tp->iss || tp->snd_nxt <= tp->seq_fin)) {
-			flags |= TH_FIN;
-			tp->seq_fin = tp->snd_lst++;
-		}
 	}
-	if (tp->snd_nxt >= tp->snd_lst)
-		return (0);
-	if (tp->tc_flags & TC_SND_URG)
-		flags |= TH_URG;
-	sent = tcp_output(tp, flags, (int)(tp->snd_lst - tp->snd_nxt), m);
-	if (!forced) {
-		tp->t_rexmt = tp->t_xmtime;
-		tp->t_rexmt_val = tp->snd_lst;
-		if ((tp->tc_flags&TC_REXMT) == 0) {
-			tp->t_rexmttl = T_REXMTTL;
-			tp->t_rtl_val = tp->snd_lst;
+	if (len > tp->t_maxseg) {
+		len = tp->t_maxseg;
+		sendalot = 1;
+	}
+	if (SEQ_LT(tp->snd_nxt + len, tp->snd_una + so->so_snd.sb_cc))
+		flags &= ~TH_FIN;
+	win = sbspace(&so->so_rcv);
+
+
+	/*
+	 * If our state indicates that FIN should be sent
+	 * and we have not yet done so, or we're retransmitting the FIN,
+	 * then we need to send.
+	 */
+	if (flags & TH_FIN &&
+	    ((tp->t_flags & TF_SENTFIN) == 0 || tp->snd_nxt == tp->snd_una))
+		goto send;
+	/*
+	 * Send if we owe peer an ACK.
+	 */
+	if (tp->t_flags & TF_ACKNOW)
+		goto send;
+	if (flags & (TH_SYN|TH_RST))
+		goto send;
+	if (SEQ_GT(tp->snd_up, tp->snd_una))
+		goto send;
+
+	/*
+	 * Sender silly window avoidance.  If connection is idle
+	 * and can send all data, a maximum segment,
+	 * at least a maximum default-size segment do it,
+	 * or are forced, do it; otherwise don't bother.
+	 * If peer's buffer is tiny, then send
+	 * when window is at least half open.
+	 * If retransmitting (possibly after persist timer forced us
+	 * to send into a small window), then must resend.
+	 */
+	if (len) {
+		if (len == tp->t_maxseg)
+			goto send;
+		if ((idle || tp->t_flags & TF_NODELAY) &&
+		    len + off >= so->so_snd.sb_cc)
+			goto send;
+		if (tp->t_force)
+			goto send;
+		if (len >= tp->max_sndwnd / 2)
+			goto send;
+		if (SEQ_LT(tp->snd_nxt, tp->snd_max))
+			goto send;
+	}
+
+	/*
+	 * Compare available window to amount of window
+	 * known to peer (as advertised window less
+	 * next expected input.)  If the difference is 35% or more of the
+	 * maximum possible window, then want to send a window update to peer.
+	 */
+	if (win > 0) {
+		int adv = win - (tp->rcv_adv - tp->rcv_nxt);
+
+		if (100 * adv / so->so_rcv.sb_hiwat >= 35)
+			goto send;
+		if (adv >= 2 * tp->t_maxseg && so->so_rcv.sb_cc == 0)
+			goto send;
+	}
+
+	/*
+	 * TCP window updates are not reliable, rather a polling protocol
+	 * using ``persist'' packets is used to insure receipt of window
+	 * updates.  The three ``states'' for the output side are:
+	 *	idle			not doing retransmits or persists
+	 *	persisting		to move a small or zero window
+	 *	(re)transmitting	and thereby not persisting
+	 *
+	 * tp->t_timer[TCPT_PERSIST]
+	 *	is set when we are in persist state.
+	 * tp->t_force
+	 *	is set when we are called to send a persist packet.
+	 * tp->t_timer[TCPT_REXMT]
+	 *	is set when we are retransmitting
+	 * The output side is idle when both timers are zero.
+	 *
+	 * If send window is too small, there is data to transmit, and no
+	 * retransmit or persist is pending, then go to persist state.
+	 * If nothing happens soon, send when timer expires:
+	 * if window is nonzero, transmit what we can,
+	 * otherwise force out a byte.
+	 */
+	if (so->so_snd.sb_cc && tp->t_timer[TCPT_REXMT] == 0 &&
+	    tp->t_timer[TCPT_PERSIST] == 0) {
+		tp->t_rxtshift = 0;
+		tcp_setpersist(tp);
+	}
+
+	/*
+	 * No reason to send a segment, just return.
+	 */
+	return (0);
+
+send:
+	/*
+	 * Grab a header mbuf, attaching a copy of data to
+	 * be transmitted, and initialize the header from
+	 * the template for sends on this connection.
+	 */
+	MGET(m, M_DONTWAIT, MT_HEADER);
+	if (m == NULL)
+		return (ENOBUFS);
+	m->m_off = MMAXOFF - sizeof (struct tcpiphdr);
+	m->m_len = sizeof (struct tcpiphdr);
+	if (len) {
+		if (tp->t_force && len == 1)
+			tcpstat.tcps_sndprobe++;
+		else if (SEQ_LT(tp->snd_nxt, tp->snd_max)) {
+			tcpstat.tcps_sndrexmitpack++;
+			tcpstat.tcps_sndrexmitbyte += len;
+		} else {
+			tcpstat.tcps_sndpack++;
+			tcpstat.tcps_sndbyte += len;
+		}
+		m->m_next = m_copy(so->so_snd.sb_mb, off, len);
+		if (m->m_next == 0)
+			len = 0;
+	} else if (tp->t_flags & TF_ACKNOW)
+		tcpstat.tcps_sndacks++;
+	else if (flags & (TH_SYN|TH_FIN|TH_RST))
+		tcpstat.tcps_sndctrl++;
+	else if (SEQ_GT(tp->snd_up, tp->snd_una))
+		tcpstat.tcps_sndurg++;
+	else
+		tcpstat.tcps_sndwinup++;
+
+	ti = mtod(m, struct tcpiphdr *);
+	if (tp->t_template == 0)
+		panic("tcp_output");
+	bcopy((caddr_t)tp->t_template, (caddr_t)ti, sizeof (struct tcpiphdr));
+
+	/*
+	 * Fill in fields, remembering maximum advertised
+	 * window for use in delaying messages about window sizes.
+	 * If resending a FIN, be sure not to use a new sequence number.
+	 */
+	if (flags & TH_FIN && tp->t_flags & TF_SENTFIN && 
+	    tp->snd_nxt == tp->snd_max)
+		tp->snd_nxt--;
+	ti->ti_seq = htonl(tp->snd_nxt);
+	ti->ti_ack = htonl(tp->rcv_nxt);
+	/*
+	 * Before ESTABLISHED, force sending of initial options
+	 * unless TCP set to not do any options.
+	 */
+	opt = NULL;
+	if (flags & TH_SYN && (tp->t_flags & TF_NOOPT) == 0) {
+		u_short mss;
+
+		mss = MIN(so->so_rcv.sb_hiwat / 2, tcp_mss(tp));
+		if (mss > IP_MSS - sizeof(struct tcpiphdr)) {
+			opt = tcp_initopt;
+			optlen = sizeof (tcp_initopt);
+			*(u_short *)(opt + 2) = htons(mss);
 		}
 	}
-	if (sent)
-		tp->snd_nxt = tp->snd_lst;
-	if ((tp->tc_flags&TC_SYN_ACKED) &&
-	    tp->snd_una > tp->t_xmt_val) {
-		tp->t_xmt = 0;
-		tp->t_xmt_val = tp->snd_lst;
+	if (opt) {
+		m0 = m->m_next;
+		m->m_next = m_get(M_DONTWAIT, MT_DATA);
+		if (m->m_next == 0) {
+			(void) m_free(m);
+			m_freem(m0);
+			return (ENOBUFS);
+		}
+		m->m_next->m_next = m0;
+		m0 = m->m_next;
+		m0->m_len = optlen;
+		bcopy((caddr_t)opt, mtod(m0, caddr_t), optlen);
+		opt = (u_char *)(mtod(m0, caddr_t) + optlen);
+		while (m0->m_len & 0x3) {
+			*opt++ = TCPOPT_EOL;
+			m0->m_len++;
+		}
+		optlen = m0->m_len;
+		ti->ti_off = (sizeof (struct tcphdr) + optlen) >> 2;
 	}
-	tp->tc_flags &= ~(TC_ACK_DUE|TC_REXMT|TC_FORCE_ONE);
-	tp->snd_hi = MAX(tp->snd_nxt, tp->snd_hi);
-	return (1);
+	ti->ti_flags = flags;
+	/*
+	 * Calculate receive window.  Don't shrink window,
+	 * but avoid silly window syndrome.
+	 */
+	if (win < so->so_rcv.sb_hiwat / 4 && win < tp->t_maxseg)
+		win = 0;
+	if (win < (int)(tp->rcv_adv - tp->rcv_nxt))
+		win = (int)(tp->rcv_adv - tp->rcv_nxt);
+	ti->ti_win = htons((u_short)win);
+	if (SEQ_GT(tp->snd_up, tp->snd_nxt)) {
+		ti->ti_urp = htons((u_short)(tp->snd_up - tp->snd_nxt));
+		ti->ti_flags |= TH_URG;
+	} else
+		/*
+		 * If no urgent pointer to send, then we pull
+		 * the urgent pointer to the left edge of the send window
+		 * so that it doesn't drift into the send window on sequence
+		 * number wraparound.
+		 */
+		tp->snd_up = tp->snd_una;		/* drag it along */
+	/*
+	 * If anything to send and we can send it all, set PUSH.
+	 * (This will keep happy those implementations which only
+	 * give data to the user when a buffer fills or a PUSH comes in.)
+	 */
+	if (len && off+len == so->so_snd.sb_cc)
+		ti->ti_flags |= TH_PUSH;
+
+	/*
+	 * Put TCP length in extended header, and then
+	 * checksum extended header and data.
+	 */
+	if (len + optlen)
+		ti->ti_len = htons((u_short)(sizeof(struct tcphdr) +
+		    optlen + len));
+	ti->ti_sum = in_cksum(m, sizeof (struct tcpiphdr) + (int)optlen + len);
+
+	/*
+	 * In transmit state, time the transmission and arrange for
+	 * the retransmit.  In persist state, just set snd_max.
+	 */
+	if (tp->t_force == 0 || tp->t_timer[TCPT_PERSIST] == 0) {
+		tcp_seq startseq = tp->snd_nxt;
+
+		/*
+		 * Advance snd_nxt over sequence space of this segment.
+		 */
+		if (flags & TH_SYN)
+			tp->snd_nxt++;
+		if (flags & TH_FIN) {
+			tp->snd_nxt++;
+			tp->t_flags |= TF_SENTFIN;
+		}
+		tp->snd_nxt += len;
+		if (SEQ_GT(tp->snd_nxt, tp->snd_max)) {
+			tp->snd_max = tp->snd_nxt;
+			/*
+			 * Time this transmission if not a retransmission and
+			 * not currently timing anything.
+			 */
+			if (tp->t_rtt == 0) {
+				tp->t_rtt = 1;
+				tp->t_rtseq = startseq;
+				tcpstat.tcps_segstimed++;
+			}
+		}
+
+		/*
+		 * Set retransmit timer if not currently set,
+		 * and not doing an ack or a keep-alive probe.
+		 * Initial value for retransmit timer is smoothed
+		 * round-trip time + 2 * round-trip time variance.
+		 * Initialize shift counter which is used for backoff
+		 * of retransmit time.
+		 */
+		if (tp->t_timer[TCPT_REXMT] == 0 &&
+		    tp->snd_nxt != tp->snd_una) {
+			tp->t_timer[TCPT_REXMT] = tp->t_rxtcur;
+			if (tp->t_timer[TCPT_PERSIST]) {
+				tp->t_timer[TCPT_PERSIST] = 0;
+				tp->t_rxtshift = 0;
+			}
+		}
+	} else
+		if (SEQ_GT(tp->snd_nxt + len, tp->snd_max))
+			tp->snd_max = tp->snd_nxt + len;
+
+	/*
+	 * Trace.
+	 */
+	if (so->so_options & SO_DEBUG)
+		tcp_trace(TA_OUTPUT, tp->t_state, tp, ti, 0);
+
+	/*
+	 * Fill in IP length and desired time to live and
+	 * send to IP level.
+	 */
+	((struct ip *)ti)->ip_len = sizeof (struct tcpiphdr) + optlen + len;
+	((struct ip *)ti)->ip_ttl = TCP_TTL;
+	error = ip_output(m, tp->t_inpcb->inp_options, &tp->t_inpcb->inp_route,
+	    so->so_options & SO_DONTROUTE);
+	if (error)
+		return (error);
+	tcpstat.tcps_sndtotal++;
+
+	/*
+	 * Data sent (as far as we can tell).
+	 * If this advertises a larger window than any other segment,
+	 * then remember the size of the advertised window.
+	 * Any pending ACK has now been sent.
+	 */
+	if (win > 0 && SEQ_GT(tp->rcv_nxt+win, tp->rcv_adv))
+		tp->rcv_adv = tp->rcv_nxt + win;
+	tp->t_flags &= ~(TF_ACKNOW|TF_DELACK);
+	if (sendalot)
+		goto again;
+	return (0);
 }
-tcp_output(tp, flags, len, dat)
+
+tcp_setpersist(tp)
 	register struct tcpcb *tp;
-	register int flags;
-	int len;
-	struct mbuf *dat;
 {
-	register struct tcpiphdr *t;			/* known to be r9 */
-	register struct mbuf *m;
-	struct socket *so = tp->t_inpcb->inp_socket;
-	register struct ip *ip;
-COUNT(TCP_OUTPUT);
-
-	if ((t = tp->t_template) == 0)
-		return (0);
-	MGET(m, 0);
-	if (m == 0)
-		return (0);
-	m->m_off = MMAXOFF - sizeof(struct tcpiphdr);
-	m->m_len = sizeof (struct tcpiphdr);
-	m->m_next = dat;
-	if (flags & TH_SYN)
-		len--;
-	if (flags & TH_FIN)
-		len--;
-	if (len < 0)
-		panic("tcp_output");
-	bcopy((caddr_t)t, mtod(m, caddr_t), sizeof (struct tcpiphdr));
-	t = mtod(m, struct tcpiphdr *);
-	if (tp->tc_flags&TC_SND_RST) {
-		flags &= ~TH_SYN;
-		flags |= TH_RST;
-	}
-	if (tp->tc_flags&TC_SYN_RCVD)
-		flags |= TH_ACK;
-	t->ti_flags = flags;
-	if (flags & TH_URG)
-		t->ti_urp = htons((u_short)tp->snd_urp);	/*XXX */
-	t->ti_win =
-	    so->so_rcv.sb_hiwat -
-		(so->so_rcv.sb_cc + tp->seqcnt);
-	if (tp->rcv_nxt + t->ti_win > tp->rcv_adv)
-		tp->rcv_adv = tp->rcv_nxt + t->ti_win;
-	if (len)
-		t->ti_len = htons((u_short)(len + sizeof (struct tcphdr)));
-	t->ti_win = htons(t->ti_win);
-	t->ti_seq = htonl(tp->snd_nxt);
-	t->ti_ackno = htonl(tp->rcv_nxt);
-	t->ti_sum = 0;		/* gratuitous? */
-	t->ti_sum = inet_cksum(m, sizeof (struct tcpiphdr) + len);
-	ip = (struct ip *)t;
-	ip->ip_v = IPVERSION;
-	ip->ip_hl = 5;
-	ip->ip_tos = 0;
-	ip->ip_len = len + sizeof(struct tcpiphdr);
-	ip->ip_id = ip_id++;
-	ip->ip_off = 0;
-	ip->ip_ttl = MAXTTL;
-	ip_send(ip);
-	return (1);
+	register t = ((tp->t_srtt >> 2) + tp->t_rttvar) >> 1;
+
+	if (tp->t_timer[TCPT_REXMT])
+		panic("tcp_output REXMT");
+	/*
+	 * Start/restart persistance timer.
+	 */
+	TCPT_RANGESET(tp->t_timer[TCPT_PERSIST],
+	    t * tcp_backoff[tp->t_rxtshift],
+	    TCPTV_PERSMIN, TCPTV_PERSMAX);
+	if (tp->t_rxtshift < TCP_MAXRXTSHIFT)
+		tp->t_rxtshift++;
 }