tcp_input.c

来自「基于组件方式开发操作系统的OSKIT源代码」· C语言 代码 · 共 2,144 行 · 第 1/4 页

				 * this is a pure ack for outstanding data.
				 */
				++tcpstat.tcps_predack;
				if ((to.to_flag & TOF_TS) != 0)
					tcp_xmit_timer(tp,
					    tcp_now - to.to_tsecr + 1);
				else if (tp->t_rtt &&
					    SEQ_GT(ti->ti_ack, tp->t_rtseq))
					tcp_xmit_timer(tp, tp->t_rtt);
				acked = ti->ti_ack - tp->snd_una;
				tcpstat.tcps_rcvackpack++;
				tcpstat.tcps_rcvackbyte += acked;
				sbdrop(&so->so_snd, acked);
				tp->snd_una = ti->ti_ack;
				m_freem(m);

				/*
				 * If all outstanding data are acked, stop
				 * retransmit timer, otherwise restart timer
				 * using current (possibly backed-off) value.
				 * If process is waiting for space,
				 * wakeup/selwakeup/signal.  If data
				 * are ready to send, let tcp_output
				 * decide between more output or persist.
				 */
				if (tp->snd_una == tp->snd_max)
					tp->t_timer[TCPT_REXMT] = 0;
				else if (tp->t_timer[TCPT_PERSIST] == 0)
					tp->t_timer[TCPT_REXMT] = tp->t_rxtcur;

				if (so->so_snd.sb_flags & SB_NOTIFY)
					sowwakeup(so);
				if (so->so_snd.sb_cc)
					(void) tcp_output(tp);
				return;
			}
		} else if (ti->ti_ack == tp->snd_una &&
		    tp->seg_next == (struct tcpiphdr *)tp &&
		    ti->ti_len <= sbspace(&so->so_rcv)) {
			/*
			 * this is a pure, in-sequence data packet
			 * with nothing on the reassembly queue and
			 * we have enough buffer space to take it.
			 */
			++tcpstat.tcps_preddat;
			tp->rcv_nxt += ti->ti_len;
			tcpstat.tcps_rcvpack++;
			tcpstat.tcps_rcvbyte += ti->ti_len;
			/*
			 * Add data to socket buffer.
			 */
			sbappend(&so->so_rcv, m);
			sorwakeup(so);
#ifdef TCP_ACK_HACK
			/*
			 * If this is a short packet, then ACK now - with Nagel
			 *	congestion avoidance sender won't send more until
			 *	he gets an ACK.
			 */
			if (tiflags & TH_PUSH) {
				tp->t_flags |= TF_ACKNOW;
				tcp_output(tp);
			} else {
				tp->t_flags |= TF_DELACK;
			}
#else
			tp->t_flags |= TF_DELACK;
#endif
			return;
		}
	}

	/*
	 * Calculate amount of space in receive window,
	 * and then do TCP input processing.
	 * Receive window is amount of space in rcv queue,
	 * but not less than advertised window.
	 */
	{ int win;

	win = sbspace(&so->so_rcv);
	if (win < 0)
		win = 0;
	tp->rcv_wnd = max(win, (int)(tp->rcv_adv - tp->rcv_nxt));
	}

	switch (tp->t_state) {

	/*
	 * If the state is LISTEN then ignore segment if it contains an RST.
	 * If the segment contains an ACK then it is bad and send a RST.
	 * If it does not contain a SYN then it is not interesting; drop it.
	 * Don't bother responding if the destination was a broadcast.
	 * Otherwise initialize tp->rcv_nxt, and tp->irs, select an initial
	 * tp->iss, and send a segment:
	 *     <SEQ=ISS><ACK=RCV_NXT><CTL=SYN,ACK>
	 * Also initialize tp->snd_nxt to tp->iss+1 and tp->snd_una to tp->iss.
	 * Fill in remote peer address fields if not previously specified.
	 * Enter SYN_RECEIVED state, and process any other fields of this
	 * segment in this state.
	 */
	case TCPS_LISTEN: {
		struct mbuf *am;
		register struct sockaddr_in *sin;

		if (tiflags & TH_RST)
			goto drop;
		if (tiflags & TH_ACK)
			goto dropwithreset;
		if ((tiflags & TH_SYN) == 0)
			goto drop;
		/*
		 * RFC1122 4.2.3.10, p. 104: discard bcast/mcast SYN
		 * in_broadcast() should never return true on a received
		 * packet with M_BCAST not set.
		 */
		if (m->m_flags & (M_BCAST|M_MCAST) ||
		    IN_MULTICAST(ntohl(ti->ti_dst.s_addr)))
			goto drop;
		am = m_get(M_DONTWAIT, MT_SONAME);	/* XXX */
		if (am == NULL)
			goto drop;
		am->m_len = sizeof (struct sockaddr_in);
		sin = mtod(am, struct sockaddr_in *);
		sin->sin_family = AF_INET;
		sin->sin_len = sizeof(*sin);
		sin->sin_addr = ti->ti_src;
		sin->sin_port = ti->ti_sport;
		bzero((caddr_t)sin->sin_zero, sizeof(sin->sin_zero));
		laddr = inp->inp_laddr;
		if (inp->inp_laddr.s_addr == INADDR_ANY)
			inp->inp_laddr = ti->ti_dst;
		if (in_pcbconnect(inp, am)) {
			inp->inp_laddr = laddr;
			(void) m_free(am);
			goto drop;
		}
		(void) m_free(am);
		tp->t_template = tcp_template(tp);
		if (tp->t_template == 0) {
			tp = tcp_drop(tp, ENOBUFS);
			dropsocket = 0;		/* socket is already gone */
			goto drop;
		}
		if ((taop = tcp_gettaocache(inp)) == NULL) {
			taop = &tao_noncached;
			bzero(taop, sizeof(*taop));
		}
		tcp_dooptions(tp, optp, optlen, ti, &to);
		if (iss)
			tp->iss = iss;
		else
			tp->iss = tcp_iss;
		tcp_iss += TCP_ISSINCR/2;
		tp->irs = ti->ti_seq;
		tcp_sendseqinit(tp);
		tcp_rcvseqinit(tp);
		/*
		 * Initialization of the tcpcb for transaction;
		 *   set SND.WND = SEG.WND,
		 *   initialize CCsend and CCrecv.
		 */
		tp->snd_wnd = tiwin;	/* initial send-window */
		tp->cc_send = CC_INC(tcp_ccgen);
		tp->cc_recv = to.to_cc;
		/*
		 * Perform TAO test on incoming CC (SEG.CC) option, if any.
		 * - compare SEG.CC against cached CC from the same host,
		 *	if any.
		 * - if SEG.CC > chached value, SYN must be new and is accepted
		 *	immediately: save new CC in the cache, mark the socket
		 *	connected, enter ESTABLISHED state, turn on flag to
		 *	send a SYN in the next segment.
		 *	A virtual advertised window is set in rcv_adv to
		 *	initialize SWS prevention.  Then enter normal segment
		 *	processing: drop SYN, process data and FIN.
		 * - otherwise do a normal 3-way handshake.
		 */
		if ((to.to_flag & TOF_CC) != 0) {
		    if (taop->tao_cc != 0 && CC_GT(to.to_cc, taop->tao_cc)) {
			taop->tao_cc = to.to_cc;
			tp->t_state = TCPS_ESTABLISHED;

			/*
			 * If there is a FIN, or if there is data and the
			 * connection is local, then delay SYN,ACK(SYN) in
			 * the hope of piggy-backing it on a response
			 * segment.  Otherwise must send ACK now in case
			 * the other side is slow starting.
			 */
			if ((tiflags & TH_FIN) || (ti->ti_len != 0 &&
			    in_localaddr(inp->inp_faddr)))
				tp->t_flags |= (TF_DELACK | TF_NEEDSYN);
			else
				tp->t_flags |= (TF_ACKNOW | TF_NEEDSYN);

			/*
			 * Limit the `virtual advertised window' to TCP_MAXWIN
			 * here.  Even if we requested window scaling, it will
			 * become effective only later when our SYN is acked.
			 */
			tp->rcv_adv += min(tp->rcv_wnd, TCP_MAXWIN);
			tcpstat.tcps_connects++;
			soisconnected(so);
			tp->t_timer[TCPT_KEEP] = tcp_keepinit;
			dropsocket = 0;		/* committed to socket */
			tcpstat.tcps_accepts++;
			goto trimthenstep6;
		    }
		/* else do standard 3-way handshake */
		} else {
		    /*
		     * No CC option, but maybe CC.NEW:
		     *   invalidate cached value.
		     */
		     taop->tao_cc = 0;
		}
		/*
		 * TAO test failed or there was no CC option,
		 *    do a standard 3-way handshake.
		 */
		tp->t_flags |= TF_ACKNOW;
		tp->t_state = TCPS_SYN_RECEIVED;
		tp->t_timer[TCPT_KEEP] = tcp_keepinit;
		dropsocket = 0;		/* committed to socket */
		tcpstat.tcps_accepts++;
		goto trimthenstep6;
		}

	/*
	 * If the state is SYN_RECEIVED:
	 *	do just the ack and RST checks from SYN_SENT state.
	 * If the state is SYN_SENT:
	 *	if seg contains an ACK, but not for our SYN, drop the input.
	 *	if seg contains a RST, then drop the connection.
	 *	if seg does not contain SYN, then drop it.
	 * Otherwise this is an acceptable SYN segment
	 *	initialize tp->rcv_nxt and tp->irs
	 *	if seg contains ack then advance tp->snd_una
	 *	if SYN has been acked change to ESTABLISHED else SYN_RCVD state
	 *	arrange for segment to be acked (eventually)
	 *	continue processing rest of data/controls, beginning with URG
	 */
	case TCPS_SYN_RECEIVED:
	case TCPS_SYN_SENT:
		if ((taop = tcp_gettaocache(inp)) == NULL) {
			taop = &tao_noncached;
			bzero(taop, sizeof(*taop));
		}

		if ((tiflags & TH_ACK) &&
		    (SEQ_LEQ(ti->ti_ack, tp->iss) ||
		     SEQ_GT(ti->ti_ack, tp->snd_max))) {
			/*
			 * If we have a cached CCsent for the remote host,
			 * hence we haven't just crashed and restarted,
			 * do not send a RST.  This may be a retransmission
			 * from the other side after our earlier ACK was lost.
			 * Our new SYN, when it arrives, will serve as the
			 * needed ACK.
			 */
			if (taop->tao_ccsent != 0)
				goto drop;
			else
				goto dropwithreset;
		}
		if (tiflags & TH_RST) {
			if (tiflags & TH_ACK)
				tp = tcp_drop(tp, ECONNREFUSED);
			goto drop;
		}
		if (tp->t_state == TCPS_SYN_RECEIVED)
			break;
		if ((tiflags & TH_SYN) == 0)
			goto drop;
		tp->snd_wnd = ti->ti_win;	/* initial send window */
		tp->cc_recv = to.to_cc;		/* foreign CC */

		tp->irs = ti->ti_seq;
		tcp_rcvseqinit(tp);
		if (tiflags & TH_ACK) {
			/*
			 * Our SYN was acked.  If segment contains CC.ECHO
			 * option, check it to make sure this segment really
			 * matches our SYN.  If not, just drop it as old
			 * duplicate, but send an RST if we're still playing
			 * by the old rules.
			 */
			if ((to.to_flag & TOF_CCECHO) &&
			    tp->cc_send != to.to_ccecho) {
				if (taop->tao_ccsent != 0)
					goto drop;
				else
					goto dropwithreset;
			}
			tcpstat.tcps_connects++;
			soisconnected(so);
			/* Do window scaling on this connection? */
			if ((tp->t_flags & (TF_RCVD_SCALE|TF_REQ_SCALE)) ==
				(TF_RCVD_SCALE|TF_REQ_SCALE)) {
				tp->snd_scale = tp->requested_s_scale;
				tp->rcv_scale = tp->request_r_scale;
			}
			/* Segment is acceptable, update cache if undefined. */
			if (taop->tao_ccsent == 0)
				taop->tao_ccsent = to.to_ccecho;

			tp->rcv_adv += tp->rcv_wnd;
			tp->snd_una++;		/* SYN is acked */
			/*
			 * If there's data, delay ACK; if there's also a FIN
			 * ACKNOW will be turned on later.
			 */
			if (ti->ti_len != 0)
				tp->t_flags |= TF_DELACK;
			else
				tp->t_flags |= TF_ACKNOW;
			/*
			 * Received <SYN,ACK> in SYN_SENT[*] state.
			 * Transitions:
			 *	SYN_SENT  --> ESTABLISHED
			 *	SYN_SENT* --> FIN_WAIT_1
			 */
			if (tp->t_flags & TF_NEEDFIN) {
				tp->t_state = TCPS_FIN_WAIT_1;
				tp->t_flags &= ~TF_NEEDFIN;
				tiflags &= ~TH_SYN;
			} else {
				tp->t_state = TCPS_ESTABLISHED;
				tp->t_timer[TCPT_KEEP] = tcp_keepidle;
			}
		} else {
		/*
		 *  Received initial SYN in SYN-SENT[*] state => simul-
		 *  taneous open.  If segment contains CC option and there is
		 *  a cached CC, apply TAO test; if it succeeds, connection is
		 *  half-synchronized.  Otherwise, do 3-way handshake:
		 *        SYN-SENT -> SYN-RECEIVED
		 *        SYN-SENT* -> SYN-RECEIVED*
		 *  If there was no CC option, clear cached CC value.
		 */
			tp->t_flags |= TF_ACKNOW;
			tp->t_timer[TCPT_REXMT] = 0;
			if (to.to_flag & TOF_CC) {
				if (taop->tao_cc != 0 &&
				    CC_GT(to.to_cc, taop->tao_cc)) {
					/*
					 * update cache and make transition:
					 *        SYN-SENT -> ESTABLISHED*
					 *        SYN-SENT* -> FIN-WAIT-1*
					 */
					taop->tao_cc = to.to_cc;
					if (tp->t_flags & TF_NEEDFIN) {
						tp->t_state = TCPS_FIN_WAIT_1;
						tp->t_flags &= ~TF_NEEDFIN;
					} else {
						tp->t_state = TCPS_ESTABLISHED;
						tp->t_timer[TCPT_KEEP] = tcp_keepidle;
					}
					tp->t_flags |= TF_NEEDSYN;
				} else
					tp->t_state = TCPS_SYN_RECEIVED;
			} else {
				/* CC.NEW or no option => invalidate cache */
				taop->tao_cc = 0;
				tp->t_state = TCPS_SYN_RECEIVED;
			}
		}

trimthenstep6:
		/*
		 * Advance ti->ti_seq to correspond to first data byte.
		 * If data, trim to stay within window,
		 * dropping FIN if necessary.
		 */
		ti->ti_seq++;
		if (ti->ti_len > tp->rcv_wnd) {
			todrop = ti->ti_len - tp->rcv_wnd;
			m_adj(m, -todrop);
			ti->ti_len = tp->rcv_wnd;
			tiflags &= ~TH_FIN;
			tcpstat.tcps_rcvpackafterwin++;
			tcpstat.tcps_rcvbyteafterwin += todrop;
		}
		tp->snd_wl1 = ti->ti_seq - 1;
		tp->rcv_up = ti->ti_seq;
		/*
		 *  Client side of transaction: already sent SYN and data.
		 *  If the remote host used T/TCP to validate the SYN,
		 *  our data will be ACK'd; if so, enter normal data segment
		 *  processing in the middle of step 5, ack processing.
		 *  Otherwise, goto step 6.
		 */
 		if (tiflags & TH_ACK)
			goto process_ACK;
		goto step6;
	/*
	 * If the state is LAST_ACK or CLOSING or TIME_WAIT:
	 *	if segment contains a SYN and CC [not CC.NEW] option:
	 *              if state == TIME_WAIT and connection duration > MSL,
	 *                  drop packet and send RST;
	 *
	 *		if SEG.CC > CCrecv then is new SYN, and can implicitly
	 *		    ack the FIN (and data) in retransmission queue.
	 *                  Complete close and delete TCPCB.  Then reprocess
	 *                  segment, hoping to find new TCPCB in LISTEN state;
	 *
	 *		else must be old SYN; drop it.
	 *      else do normal processing.
	 */
	case TCPS_LAST_ACK:
	case TCPS_CLOSING:
	case TCPS_TIME_WAIT:
		if ((tiflags & TH_SYN) &&
		    (to.to_flag & TOF_CC) && tp->cc_recv != 0) {
			if (tp->t_state == TCPS_TIME_WAIT &&
					tp->t_duration > TCPTV_MSL)
				goto dropwithreset;
			if (CC_GT(to.to_cc, tp->cc_recv)) {
				tp = tcp_close(tp);
				goto findpcb;
			}
			else
				goto drop;
		}
 		break;  /* continue normal processing */
	}

	/*
	 * States other than LISTEN or SYN_SENT.
	 * First check timestamp, if present.
	 * Then check the connection count, if present.
	 * Then check that at least some bytes of segment are within
	 * receive window.  If segment begins before rcv_nxt,
	 * drop leading data (and SYN); if nothing left, just ack.
	 *
	 * RFC 1323 PAWS: If we have a timestamp reply on this segment
	 * and it's less than ts_recent, drop it.
	 */
	if ((to.to_flag & TOF_TS) != 0 && (tiflags & TH_RST) == 0 &&
	    tp->ts_recent && TSTMP_LT(to.to_tsval, tp->ts_recent)) {

		/* Check to see if ts_recent is over 24 days old.  */
		if ((int)(tcp_now - tp->ts_recent_age) > TCP_PAWS_IDLE) {
			/*
			 * Invalidate ts_recent.  If this segment updates
			 * ts_recent, the age will be reset later and ts_recent
			 * will get a valid value.  If it does not, setting
			 * ts_recent to zero will at least satisfy the
			 * requirement that zero be placed in the timestamp
			 * echo reply when ts_recent isn't valid.  The
			 * age isn't reset until we get a valid ts_recent
			 * because we don't want out-of-order segments to be
			 * dropped when ts_recent is old.
			 */
			tp->ts_recent = 0;
		} else {
			tcpstat.tcps_rcvduppack++;
			tcpstat.tcps_rcvdupbyte += ti->ti_len;
			tcpstat.tcps_pawsdrop++;
			goto dropafterack;
		}
	}

	/*
	 * T/TCP mechanism
	 *   If T/TCP was negotiated and the segment doesn't have CC,
	 *   or if it's CC is wrong then drop the segment.
	 *   RST segments do not have to comply with this.
	 */
	if ((tp->t_flags & (TF_REQ_CC|TF_RCVD_CC)) == (TF_REQ_CC|TF_RCVD_CC) &&
	    ((to.to_flag & TOF_CC) == 0 || tp->cc_recv != to.to_cc) &&
	    (tiflags & TH_RST) == 0)
 		goto dropafterack;

	todrop = tp->rcv_nxt - ti->ti_seq;
	if (todrop > 0) {
		if (tiflags & TH_SYN) {
			tiflags &= ~TH_SYN;
			ti->ti_seq++;
			if (ti->ti_urp > 1)
				ti->ti_urp--;
			else
				tiflags &= ~TH_URG;
			todrop--;
		}
		/*
		 * Following if statement from Stevens, vol. 2, p. 960.
		 */
		if (todrop > ti->ti_len
		    || (todrop == ti->ti_len && (tiflags & TH_FIN) == 0)) {
			/*
			 * Any valid FIN must be to the left of the window.
			 * At this point the FIN must be a duplicate or out
			 * of sequence; drop it.
			 */
			tiflags &= ~TH_FIN;

			/*
			 * Send an ACK to resynchronize and drop any data.
			 * But keep on processing for RST or ACK.
			 */
			tp->t_flags |= TF_ACKNOW;
			todrop = ti->ti_len;
			tcpstat.tcps_rcvduppack++;
			tcpstat.tcps_rcvdupbyte += todrop;
		} else {
			tcpstat.tcps_rcvpartduppack++;
			tcpstat.tcps_rcvpartdupbyte += todrop;
		}
		m_adj(m, todrop);
		ti->ti_seq += todrop;
		ti->ti_len -= todrop;
		if (ti->ti_urp > todrop)
			ti->ti_urp -= todrop;
		else {
			tiflags &= ~TH_URG;
			ti->ti_urp = 0;
		}
	}

	/*
	 * If new data are received on a connection after the
	 * user processes are gone, then RST the other end.
	 */
	if ((so->so_state & SS_NOFDREF) &&
	    tp->t_state > TCPS_CLOSE_WAIT && ti->ti_len) {
		tp = tcp_close(tp);
		tcpstat.tcps_rcvafterclose++;
		goto dropwithreset;
	}

	/*
	 * If segment ends after window, drop trailing data
	 * (and PUSH and FIN); if nothing left, just ACK.
	 */