tcp_input.c

ReactOS是一些高手根据Windows XP的内核编写出的类XP。内核实现机理和API函数调用几乎相同。甚至可以兼容XP的程序。喜欢研究系统内核的人可以看一看。

tcp_dooptions(tp, cp, cnt, ti, to)
	struct tcpcb *tp;
	u_char *cp;
	int cnt;
	struct tcpiphdr *ti;
	struct tcpopt *to;
{
	u_short mss = 0;
	int opt, optlen;

	for (; cnt > 0; cnt -= optlen, cp += optlen) {
		opt = cp[0];
		if (opt == TCPOPT_EOL)
			break;
		if (opt == TCPOPT_NOP)
			optlen = 1;
		else {
			optlen = cp[1];
			if (optlen <= 0)
				break;
		}
		switch (opt) {

		default:
			continue;

		case TCPOPT_MAXSEG:
			if (optlen != TCPOLEN_MAXSEG)
				continue;
			if (!(ti->ti_flags & TH_SYN))
				continue;
			bcopy((char *) cp + 2, (char *) &mss, sizeof(mss));
			NTOHS(mss);
			break;

		case TCPOPT_WINDOW:
			if (optlen != TCPOLEN_WINDOW)
				continue;
			if (!(ti->ti_flags & TH_SYN))
				continue;
			tp->t_flags |= TF_RCVD_SCALE;
			tp->requested_s_scale = min(cp[2], TCP_MAX_WINSHIFT);
			break;

		case TCPOPT_TIMESTAMP:
			if (optlen != TCPOLEN_TIMESTAMP)
				continue;
			to->to_flag |= TOF_TS;
			bcopy((char *)cp + 2,
			    (char *)&to->to_tsval, sizeof(to->to_tsval));
			NTOHL(to->to_tsval);
			bcopy((char *)cp + 6,
			    (char *)&to->to_tsecr, sizeof(to->to_tsecr));
			NTOHL(to->to_tsecr);

			/*
			 * A timestamp received in a SYN makes
			 * it ok to send timestamp requests and replies.
			 */
			if (ti->ti_flags & TH_SYN) {
				tp->t_flags |= TF_RCVD_TSTMP;
				tp->ts_recent = to->to_tsval;
				tp->ts_recent_age = tcp_now;
			}
			break;
		case TCPOPT_CC:
			if (optlen != TCPOLEN_CC)
				continue;
			to->to_flag |= TOF_CC;
			bcopy((char *)cp + 2,
			    (char *)&to->to_cc, sizeof(to->to_cc));
			NTOHL(to->to_cc);
			/*
			 * A CC or CC.new option received in a SYN makes
			 * it ok to send CC in subsequent segments.
			 */
			if (ti->ti_flags & TH_SYN)
				tp->t_flags |= TF_RCVD_CC;
			break;
		case TCPOPT_CCNEW:
			if (optlen != TCPOLEN_CC)
				continue;
			if (!(ti->ti_flags & TH_SYN))
				continue;
			to->to_flag |= TOF_CCNEW;
			bcopy((char *)cp + 2,
			    (char *)&to->to_cc, sizeof(to->to_cc));
			NTOHL(to->to_cc);
			/*
			 * A CC or CC.new option received in a SYN makes
			 * it ok to send CC in subsequent segments.
			 */
			tp->t_flags |= TF_RCVD_CC;
			break;
		case TCPOPT_CCECHO:
			if (optlen != TCPOLEN_CC)
				continue;
			if (!(ti->ti_flags & TH_SYN))
				continue;
			to->to_flag |= TOF_CCECHO;
			bcopy((char *)cp + 2,
			    (char *)&to->to_ccecho, sizeof(to->to_ccecho));
			NTOHL(to->to_ccecho);
			break;
		}
	}
	if (ti->ti_flags & TH_SYN)
		tcp_mss(tp, mss);	/* sets t_maxseg */
}

/*
 * Pull out of band byte out of a segment so
 * it doesn't appear in the user's data queue.
 * It is still reflected in the segment length for
 * sequencing purposes.
 */
void
tcp_pulloutofband(so, ti, m)
	struct socket *so;
	struct tcpiphdr *ti;
	register struct mbuf *m;
{
	int cnt = ti->ti_urp - 1;

	while (cnt >= 0) {
		if (m->m_len > cnt) {
			char *cp = mtod(m, caddr_t) + cnt;
			struct tcpcb *tp = sototcpcb(so);

			tp->t_iobc = *cp;
			tp->t_oobflags |= TCPOOB_HAVEDATA;
			bcopy(cp+1, cp, (unsigned)(m->m_len - cnt - 1));
			m->m_len--;
			return;
		}
		cnt -= m->m_len;
		m = m->m_next;
		if (m == 0)
			break;
	}
	panic("tcp_pulloutofband");
}

/*
 * Collect new round-trip time estimate
 * and update averages and current timeout.
 */
void
tcp_xmit_timer(tp, rtt)
	register struct tcpcb *tp;
	short rtt;
{
	register short delta;

	tcpstat.tcps_rttupdated++;
	tp->t_rttupdated++;
	if (tp->t_srtt != 0) {
		/*
		 * srtt is stored as fixed point with 3 bits after the
		 * binary point (i.e., scaled by 8).  The following magic
		 * is equivalent to the smoothing algorithm in rfc793 with
		 * an alpha of .875 (srtt = rtt/8 + srtt*7/8 in fixed
		 * point).  Adjust rtt to origin 0.
		 */
		delta = rtt - 1 - (tp->t_srtt >> TCP_RTT_SHIFT);
		if ((tp->t_srtt += delta) <= 0)
			tp->t_srtt = 1;
		/*
		 * We accumulate a smoothed rtt variance (actually, a
		 * smoothed mean difference), then set the retransmit
		 * timer to smoothed rtt + 4 times the smoothed variance.
		 * rttvar is stored as fixed point with 2 bits after the
		 * binary point (scaled by 4).  The following is
		 * equivalent to rfc793 smoothing with an alpha of .75
		 * (rttvar = rttvar*3/4 + |delta| / 4).  This replaces
		 * rfc793's wired-in beta.
		 */
		if (delta < 0)
			delta = -delta;
		delta -= (tp->t_rttvar >> TCP_RTTVAR_SHIFT);
		if ((tp->t_rttvar += delta) <= 0)
			tp->t_rttvar = 1;
	} else {
		/*
		 * No rtt measurement yet - use the unsmoothed rtt.
		 * Set the variance to half the rtt (so our first
		 * retransmit happens at 3*rtt).
		 */
		tp->t_srtt = rtt << TCP_RTT_SHIFT;
		tp->t_rttvar = rtt << (TCP_RTTVAR_SHIFT - 1);
	}
	tp->t_rtt = 0;
	tp->t_rxtshift = 0;

	/*
	 * the retransmit should happen at rtt + 4 * rttvar.
	 * Because of the way we do the smoothing, srtt and rttvar
	 * will each average +1/2 tick of bias.  When we compute
	 * the retransmit timer, we want 1/2 tick of rounding and
	 * 1 extra tick because of +-1/2 tick uncertainty in the
	 * firing of the timer.  The bias will give us exactly the
	 * 1.5 tick we need.  But, because the bias is
	 * statistical, we have to test that we don't drop below
	 * the minimum feasible timer (which is 2 ticks).
	 */
	TCPT_RANGESET(tp->t_rxtcur, TCP_REXMTVAL(tp),
	    tp->t_rttmin, TCPTV_REXMTMAX);

	/*
	 * We received an ack for a packet that wasn't retransmitted;
	 * it is probably safe to discard any error indications we've
	 * received recently.  This isn't quite right, but close enough
	 * for now (a route might have failed after we sent a segment,
	 * and the return path might not be symmetrical).
	 */
	tp->t_softerror = 0;
}

/*
 * Determine a reasonable value for maxseg size.
 * If the route is known, check route for mtu.
 * If none, use an mss that can be handled on the outgoing
 * interface without forcing IP to fragment; if bigger than
 * an mbuf cluster (MCLBYTES), round down to nearest multiple of MCLBYTES
 * to utilize large mbufs.  If no route is found, route has no mtu,
 * or the destination isn't local, use a default, hopefully conservative
 * size (usually 512 or the default IP max size, but no more than the mtu
 * of the interface), as we can't discover anything about intervening
 * gateways or networks.  We also initialize the congestion/slow start
 * window to be a single segment if the destination isn't local.
 * While looking at the routing entry, we also initialize other path-dependent
 * parameters from pre-set or cached values in the routing entry.
 *
 * Also take into account the space needed for options that we
 * send regularly.  Make maxseg shorter by that amount to assure
 * that we can send maxseg amount of data even when the options
 * are present.  Store the upper limit of the length of options plus
 * data in maxopd.
 *
 * NOTE that this routine is only called when we process an incoming
 * segment, for outgoing segments only tcp_mssopt is called.
 *
 * In case of T/TCP, we call this routine during implicit connection
 * setup as well (offer = -1), to initialize maxseg from the cached
 * MSS of our peer.
 */
void
tcp_mss(tp, offer)
	struct tcpcb *tp;
	int offer;
{
	register struct rtentry *rt;
	struct ifnet *ifp;
	register int rtt, mss;
	u_long bufsize;
	struct inpcb *inp;
	struct socket *so;
	struct rmxp_tao *taop;
	int origoffer = offer;

	inp = tp->t_inpcb;
	if ((rt = tcp_rtlookup(inp)) == NULL) {
		tp->t_maxopd = tp->t_maxseg = tcp_mssdflt;
		return;
	}
#ifndef __REACTOS__
	ifp = rt->rt_ifp;
#endif
	so = inp->inp_socket;

	taop = rmx_taop(rt->rt_rmx);
	/*
	 * Offer == -1 means that we didn't receive SYN yet,
	 * use cached value in that case;
	 */
	if (offer == -1)
		offer = taop->tao_mssopt;
	/*
	 * Offer == 0 means that there was no MSS on the SYN segment,
	 * in this case we use tcp_mssdflt.
	 */
	if (offer == 0)
		offer = tcp_mssdflt;
	else
		/*
		 * Sanity check: make sure that maxopd will be large
		 * enough to allow some data on segments even is the
		 * all the option space is used (40bytes).  Otherwise
		 * funny things may happen in tcp_output.
		 */
		offer = max(offer, 64);
	taop->tao_mssopt = offer;

	/*
	 * While we're here, check if there's an initial rtt
	 * or rttvar.  Convert from the route-table units
	 * to scaled multiples of the slow timeout timer.
	 */
	if (tp->t_srtt == 0 && (rtt = rt->rt_rmx.rmx_rtt)) {
		/*
		 * XXX the lock bit for RTT indicates that the value
		 * is also a minimum value; this is subject to time.
		 */
		if (rt->rt_rmx.rmx_locks & RTV_RTT)
			tp->t_rttmin = rtt / (RTM_RTTUNIT / PR_SLOWHZ);
		tp->t_srtt = rtt / (RTM_RTTUNIT / (PR_SLOWHZ * TCP_RTT_SCALE));
		tcpstat.tcps_usedrtt++;
		if (rt->rt_rmx.rmx_rttvar) {
			tp->t_rttvar = rt->rt_rmx.rmx_rttvar /
			    (RTM_RTTUNIT / (PR_SLOWHZ * TCP_RTTVAR_SCALE));
			tcpstat.tcps_usedrttvar++;
		} else {
			/* default variation is +- 1 rtt */
			tp->t_rttvar =
			    tp->t_srtt * TCP_RTTVAR_SCALE / TCP_RTT_SCALE;
		}
		TCPT_RANGESET(tp->t_rxtcur,
		    ((tp->t_srtt >> 2) + tp->t_rttvar) >> 1,
		    tp->t_rttmin, TCPTV_REXMTMAX);
	}
	/*
	 * if there's an mtu associated with the route, use it
	 */
	if (rt->rt_rmx.rmx_mtu)
		mss = rt->rt_rmx.rmx_mtu - sizeof(struct tcpiphdr);
	else
	{
		mss = ifp->if_mtu - sizeof(struct tcpiphdr);
		if (!in_localaddr(inp->inp_faddr))
			mss = min(mss, tcp_mssdflt);
	}
	mss = min(mss, offer);
	/*
	 * maxopd stores the maximum length of data AND options
	 * in a segment; maxseg is the amount of data in a normal
	 * segment.  We need to store this value (maxopd) apart
	 * from maxseg, because now every segment carries options
	 * and thus we normally have somewhat less data in segments.
	 */
	tp->t_maxopd = mss;

	/*
	 * In case of T/TCP, origoffer==-1 indicates, that no segments
	 * were received yet.  In this case we just guess, otherwise
	 * we do the same as before T/TCP.
	 */
 	if ((tp->t_flags & (TF_REQ_TSTMP|TF_NOOPT)) == TF_REQ_TSTMP &&
	    (origoffer == -1 ||
	     (tp->t_flags & TF_RCVD_TSTMP) == TF_RCVD_TSTMP))
		mss -= TCPOLEN_TSTAMP_APPA;
 	if ((tp->t_flags & (TF_REQ_CC|TF_NOOPT)) == TF_REQ_CC &&
	    (origoffer == -1 ||
	     (tp->t_flags & TF_RCVD_CC) == TF_RCVD_CC))
		mss -= TCPOLEN_CC_APPA;

#if	(MCLBYTES & (MCLBYTES - 1)) == 0
		if (mss > MCLBYTES)
			mss &= ~(MCLBYTES-1);
#else
		if (mss > MCLBYTES)
			mss = mss / MCLBYTES * MCLBYTES;
#endif
	/*
	 * If there's a pipesize, change the socket buffer
	 * to that size.  Make the socket buffers an integral
	 * number of mss units; if the mss is larger than
	 * the socket buffer, decrease the mss.
	 */
#ifdef RTV_SPIPE
	if ((bufsize = rt->rt_rmx.rmx_sendpipe) == 0)
#endif
		bufsize = so->so_snd.sb_hiwat;
	if (bufsize < mss)
		mss = bufsize;
	else {
		bufsize = roundup(bufsize, mss);
		if (bufsize > sb_max)
			bufsize = sb_max;
		(void)sbreserve(&so->so_snd, bufsize);
	}
	tp->t_maxseg = mss;

#ifdef RTV_RPIPE
	if ((bufsize = rt->rt_rmx.rmx_recvpipe) == 0)
#endif
		bufsize = so->so_rcv.sb_hiwat;
	if (bufsize > mss) {
		bufsize = roundup(bufsize, mss);
		if (bufsize > sb_max)
			bufsize = sb_max;
		(void)sbreserve(&so->so_rcv, bufsize);
	}
	/*
	 * Don't force slow-start on local network.
	 */
	if (!in_localaddr(inp->inp_faddr))
		tp->snd_cwnd = mss;

	if (rt->rt_rmx.rmx_ssthresh) {
		/*
		 * There's some sort of gateway or interface
		 * buffer limit on the path.  Use this to set
		 * the slow start threshhold, but set the
		 * threshold to no less than 2*mss.
		 */
		tp->snd_ssthresh = max(2 * mss, rt->rt_rmx.rmx_ssthresh);
		tcpstat.tcps_usedssthresh++;
	}
}

/*
 * Determine the MSS option to send on an outgoing SYN.
 */
int
tcp_mssopt(tp)
	struct tcpcb *tp;
{
	struct rtentry *rt;

	rt = tcp_rtlookup(tp->t_inpcb);
	if (rt == NULL)
		return tcp_mssdflt;
#ifndef __REACTOS__
	return rt->rt_ifp->if_mtu - sizeof(struct tcpiphdr);
#else
	return tcp_mssdflt;
#endif
}
#endif /* TUBA_INCLUDE */