d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
0b757633b3
freebsd-skq/sys/netinet/tcp_reass.c
Brian Feldman ecf723083f Implement RLIMIT_SBSIZE in the kernel. This is a per-uid sockbuf total 
usage limit.
1999-10-09 20:42:17 +00:00

2352 lines
67 KiB
C
Raw Blame History

/*
* Copyright (c) 1982, 1986, 1988, 1990, 1993, 1994, 1995
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)tcp_input.c 8.12 (Berkeley) 5/24/95
* $FreeBSD$
*/
#include "opt_ipfw.h" /* for ipfw_fwd */
#include "opt_tcpdebug.h"
#include "opt_tcp_input.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/sysctl.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/proc.h> /* for proc0 declaration */
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/syslog.h>
#include <machine/cpu.h> /* before tcp_seq.h, for tcp_random18() */
#include <net/if.h>
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_icmp.h> /* for ICMP_BANDLIM */
#include <netinet/in_pcb.h>
#include <netinet/ip_var.h>
#include <netinet/icmp_var.h> /* for ICMP_BANDLIM */
#include <netinet/tcp.h>
#include <netinet/tcp_fsm.h>
#include <netinet/tcp_seq.h>
#include <netinet/tcp_timer.h>
#include <netinet/tcp_var.h>
#include <netinet/tcpip.h>
#ifdef TCPDEBUG
#include <netinet/tcp_debug.h>
static struct tcpiphdr tcp_saveti;
#endif
static int tcprexmtthresh = 3;
tcp_seq tcp_iss;
tcp_cc tcp_ccgen;
struct tcpstat tcpstat;
SYSCTL_STRUCT(_net_inet_tcp, TCPCTL_STATS, stats, CTLFLAG_RD,
&tcpstat , tcpstat, "TCP statistics (struct tcpstat, netinet/tcp_var.h)");
static int log_in_vain = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, log_in_vain, CTLFLAG_RW,
&log_in_vain, 0, "Log all incoming TCP connections");
static int blackhole = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, blackhole, CTLFLAG_RW,
&blackhole, 0, "Do not send RST when dropping refused connections");
int tcp_delack_enabled = 1;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, delayed_ack, CTLFLAG_RW,
&tcp_delack_enabled, 0,
"Delay ACK to try and piggyback it onto a data packet");
#ifdef TCP_DROP_SYNFIN
static int drop_synfin = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, drop_synfin, CTLFLAG_RW,
&drop_synfin, 0, "Drop TCP packets with SYN+FIN set");
#endif
#ifdef TCP_RESTRICT_RST
static int restrict_rst = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, restrict_rst, CTLFLAG_RW,
&restrict_rst, 0, "Restrict RST emission");
#endif
struct inpcbhead tcb;
struct inpcbinfo tcbinfo;
static void tcp_dooptions __P((struct tcpcb *,
u_char *, int, struct tcpiphdr *, struct tcpopt *));
static void tcp_pulloutofband __P((struct socket *,
struct tcpiphdr *, struct mbuf *));
static int tcp_reass __P((struct tcpcb *, struct tcpiphdr *, struct mbuf *));
static void tcp_xmit_timer __P((struct tcpcb *, int));
/*
* Insert segment ti into reassembly queue of tcp with
* control block tp. Return TH_FIN if reassembly now includes
* a segment with FIN. The macro form does the common case inline
* (segment is the next to be received on an established connection,
* and the queue is empty), avoiding linkage into and removal
* from the queue and repetition of various conversions.
* Set DELACK for segments received in order, but ack immediately
* when segments are out of order (so fast retransmit can work).
*/
#define TCP_REASS(tp, ti, m, so, flags) { \
if ((ti)->ti_seq == (tp)->rcv_nxt && \
(tp)->t_segq == NULL && \
(tp)->t_state == TCPS_ESTABLISHED) { \
if (tcp_delack_enabled) \
callout_reset(tp->tt_delack, tcp_delacktime, \
tcp_timer_delack, tp); \
else \
tp->t_flags |= TF_ACKNOW; \
(tp)->rcv_nxt += (ti)->ti_len; \
flags = (ti)->ti_flags & TH_FIN; \
tcpstat.tcps_rcvpack++;\
tcpstat.tcps_rcvbyte += (ti)->ti_len;\
sbappend(&(so)->so_rcv, (m)); \
sorwakeup(so); \
} else { \
(flags) = tcp_reass((tp), (ti), (m)); \
tp->t_flags |= TF_ACKNOW; \
} \
}
static int
tcp_reass(tp, ti, m)
register struct tcpcb *tp;
register struct tcpiphdr *ti;
struct mbuf *m;
{
struct mbuf *q;
struct mbuf *p;
struct mbuf *nq;
struct socket *so = tp->t_inpcb->inp_socket;
int flags;
#define GETTCP(m) ((struct tcpiphdr *)m->m_pkthdr.header)
/*
* Call with ti==0 after become established to
* force pre-ESTABLISHED data up to user socket.
*/
if (ti == 0)
goto present;
m->m_pkthdr.header = ti;
/*
* Find a segment which begins after this one does.
*/
for (q = tp->t_segq, p = NULL; q; p = q, q = q->m_nextpkt)
if (SEQ_GT(GETTCP(q)->ti_seq, ti->ti_seq))
break;
/*
* If there is a preceding segment, it may provide some of
* our data already. If so, drop the data from the incoming
* segment. If it provides all of our data, drop us.
*/
if (p != NULL) {
register int i;
/* conversion to int (in i) handles seq wraparound */
i = GETTCP(p)->ti_seq + GETTCP(p)->ti_len - ti->ti_seq;
if (i > 0) {
if (i >= ti->ti_len) {
tcpstat.tcps_rcvduppack++;
tcpstat.tcps_rcvdupbyte += ti->ti_len;
m_freem(m);
/*
* Try to present any queued data
* at the left window edge to the user.
* This is needed after the 3-WHS
* completes.
*/
goto present; /* ??? */
}
m_adj(m, i);
ti->ti_len -= i;
ti->ti_seq += i;
}
}
tcpstat.tcps_rcvoopack++;
tcpstat.tcps_rcvoobyte += ti->ti_len;
/*
* While we overlap succeeding segments trim them or,
* if they are completely covered, dequeue them.
*/
while (q) {
register int i = (ti->ti_seq + ti->ti_len) - GETTCP(q)->ti_seq;
if (i <= 0)
break;
if (i < GETTCP(q)->ti_len) {
GETTCP(q)->ti_seq += i;
GETTCP(q)->ti_len -= i;
m_adj(q, i);
break;
}
nq = q->m_nextpkt;
if (p)
p->m_nextpkt = nq;
else
tp->t_segq = nq;
m_freem(q);
q = nq;
}
if (p == NULL) {
m->m_nextpkt = tp->t_segq;
tp->t_segq = m;
} else {
m->m_nextpkt = p->m_nextpkt;
p->m_nextpkt = m;
}
present:
/*
* Present data to user, advancing rcv_nxt through
* completed sequence space.
*/
if (!TCPS_HAVEESTABLISHED(tp->t_state))
return (0);
q = tp->t_segq;
if (!q || GETTCP(q)->ti_seq != tp->rcv_nxt)
return (0);
do {
tp->rcv_nxt += GETTCP(q)->ti_len;
flags = GETTCP(q)->ti_flags & TH_FIN;
nq = q->m_nextpkt;
tp->t_segq = nq;
q->m_nextpkt = NULL;
if (so->so_state & SS_CANTRCVMORE)
m_freem(q);
else
sbappend(&so->so_rcv, q);
q = nq;
} while (q && GETTCP(q)->ti_seq == tp->rcv_nxt);
sorwakeup(so);
return (flags);
#undef GETTCP
}
/*
* TCP input routine, follows pages 65-76 of the
* protocol specification dated September, 1981 very closely.
*/
void
tcp_input(m, iphlen)
register struct mbuf *m;
int iphlen;
{
register struct tcpiphdr *ti;
register struct inpcb *inp;
u_char *optp = NULL;
int optlen = 0;
int len, tlen, off;
register struct tcpcb *tp = 0;
register int tiflags;
struct socket *so = 0;
int todrop, acked, ourfinisacked, needoutput = 0;
struct in_addr laddr;
int dropsocket = 0;
int iss = 0;
u_long tiwin;
struct tcpopt to; /* options in this segment */
struct rmxp_tao *taop; /* pointer to our TAO cache entry */
struct rmxp_tao tao_noncached; /* in case there's no cached entry */
#ifdef TCPDEBUG
short ostate = 0;
#endif
bzero((char *)&to, sizeof(to));
tcpstat.tcps_rcvtotal++;
/*
* Get IP and TCP header together in first mbuf.
* Note: IP leaves IP header in first mbuf.
*/
ti = mtod(m, struct tcpiphdr *);
if (iphlen > sizeof (struct ip))
ip_stripoptions(m, (struct mbuf *)0);
if (m->m_len < sizeof (struct tcpiphdr)) {
if ((m = m_pullup(m, sizeof (struct tcpiphdr))) == 0) {
tcpstat.tcps_rcvshort++;
return;
}
ti = mtod(m, struct tcpiphdr *);
}
/*
* Checksum extended TCP header and data.
*/
tlen = ((struct ip *)ti)->ip_len;
len = sizeof (struct ip) + tlen;
bzero(ti->ti_x1, sizeof(ti->ti_x1));
ti->ti_len = (u_short)tlen;
HTONS(ti->ti_len);
ti->ti_sum = in_cksum(m, len);
if (ti->ti_sum) {
tcpstat.tcps_rcvbadsum++;
goto drop;
}
/*
* Check that TCP offset makes sense,
* pull out TCP options and adjust length. XXX
*/
off = ti->ti_off << 2;
if (off < sizeof (struct tcphdr) || off > tlen) {
tcpstat.tcps_rcvbadoff++;
goto drop;
}
tlen -= off;
ti->ti_len = tlen;
if (off > sizeof (struct tcphdr)) {
if (m->m_len < sizeof(struct ip) + off) {
if ((m = m_pullup(m, sizeof (struct ip) + off)) == 0) {
tcpstat.tcps_rcvshort++;
return;
}
ti = mtod(m, struct tcpiphdr *);
}
optlen = off - sizeof (struct tcphdr);
optp = mtod(m, u_char *) + sizeof (struct tcpiphdr);
}
tiflags = ti->ti_flags;
#ifdef TCP_DROP_SYNFIN
/*
* If the drop_synfin option is enabled, drop all packets with
* both the SYN and FIN bits set. This prevents e.g. nmap from
* identifying the TCP/IP stack.
*
* This is incompatible with RFC1644 extensions (T/TCP).
*/
if (drop_synfin && (tiflags & (TH_SYN|TH_FIN)) == (TH_SYN|TH_FIN))
goto drop;
#endif
/*
* Convert TCP protocol specific fields to host format.
*/
NTOHL(ti->ti_seq);
NTOHL(ti->ti_ack);
NTOHS(ti->ti_win);
NTOHS(ti->ti_urp);
/*
* Drop TCP, IP headers and TCP options.
*/
m->m_data += sizeof(struct tcpiphdr)+off-sizeof(struct tcphdr);
m->m_len -= sizeof(struct tcpiphdr)+off-sizeof(struct tcphdr);
/*
* Locate pcb for segment.
*/
findpcb:
#ifdef IPFIREWALL_FORWARD
if (ip_fw_fwd_addr != NULL) {
/*
* Diverted. Pretend to be the destination.
* already got one like this?
*/
inp = in_pcblookup_hash(&tcbinfo, ti->ti_src, ti->ti_sport,
ti->ti_dst, ti->ti_dport, 0);
if (!inp) {
/*
* No, then it's new. Try find the ambushing socket
*/
if (!ip_fw_fwd_addr->sin_port) {
inp = in_pcblookup_hash(&tcbinfo, ti->ti_src,
ti->ti_sport, ip_fw_fwd_addr->sin_addr,
ti->ti_dport, 1);
} else {
inp = in_pcblookup_hash(&tcbinfo,
ti->ti_src, ti->ti_sport,
ip_fw_fwd_addr->sin_addr,
ntohs(ip_fw_fwd_addr->sin_port), 1);
}
}
ip_fw_fwd_addr = NULL;
} else
#endif /* IPFIREWALL_FORWARD */
inp = in_pcblookup_hash(&tcbinfo, ti->ti_src, ti->ti_sport,
ti->ti_dst, ti->ti_dport, 1);
/*
* If the state is CLOSED (i.e., TCB does not exist) then
* all data in the incoming segment is discarded.
* If the TCB exists but is in CLOSED state, it is embryonic,
* but should either do a listen or a connect soon.
*/
if (inp == NULL) {
if (log_in_vain) {
char buf[4*sizeof "123"];
strcpy(buf, inet_ntoa(ti->ti_dst));
switch (log_in_vain) {
case 1:
if(tiflags & TH_SYN)
log(LOG_INFO,
"Connection attempt to TCP %s:%d from %s:%d\n",
buf, ntohs(ti->ti_dport),
inet_ntoa(ti->ti_src),
ntohs(ti->ti_sport));
break;
case 2:
log(LOG_INFO,
"Connection attempt to TCP %s:%d from %s:%d flags:0x%x\n",
buf, ntohs(ti->ti_dport), inet_ntoa(ti->ti_src),
ntohs(ti->ti_sport), tiflags);
break;
default:
break;
}
}
#ifdef ICMP_BANDLIM
if (badport_bandlim(1) < 0)
goto drop;
#endif
if (blackhole) {
switch (blackhole) {
case 1:
if (tiflags & TH_SYN)
goto drop;
break;
case 2:
goto drop;
default:
goto drop;
}
}
goto dropwithreset;
}
tp = intotcpcb(inp);
if (tp == 0)
goto dropwithreset;
if (tp->t_state == TCPS_CLOSED)
goto drop;
/* Unscale the window into a 32-bit value. */
if ((tiflags & TH_SYN) == 0)
tiwin = ti->ti_win << tp->snd_scale;
else
tiwin = ti->ti_win;
so = inp->inp_socket;
if (so->so_options & (SO_DEBUG|SO_ACCEPTCONN)) {
#ifdef TCPDEBUG
if (so->so_options & SO_DEBUG) {
ostate = tp->t_state;
tcp_saveti = *ti;
}
#endif
if (so->so_options & SO_ACCEPTCONN) {
register struct tcpcb *tp0 = tp;
struct socket *so2;
if ((tiflags & (TH_RST|TH_ACK|TH_SYN)) != TH_SYN) {
/*
* Note: dropwithreset makes sure we don't
* send a RST in response to a RST.
*/
if (tiflags & TH_ACK) {
tcpstat.tcps_badsyn++;
goto dropwithreset;
}
goto drop;
}
so2 = sonewconn(so, 0);
if (so2 == 0) {
tcpstat.tcps_listendrop++;
so2 = sodropablereq(so);
if (so2) {
tcp_drop(sototcpcb(so2), ETIMEDOUT);
so2 = sonewconn(so, 0);
}
if (!so2)
goto drop;
}
so = so2;
/*
* This is ugly, but ....
*
* Mark socket as temporary until we're
* committed to keeping it. The code at
* ``drop'' and ``dropwithreset'' check the
* flag dropsocket to see if the temporary
* socket created here should be discarded.
* We mark the socket as discardable until
* we're committed to it below in TCPS_LISTEN.
*/
dropsocket++;
inp = (struct inpcb *)so->so_pcb;
inp->inp_laddr = ti->ti_dst;
inp->inp_lport = ti->ti_dport;
if (in_pcbinshash(inp) != 0) {
/*
* Undo the assignments above if we failed to
* put the PCB on the hash lists.
*/
inp->inp_laddr.s_addr = INADDR_ANY;
inp->inp_lport = 0;
goto drop;
}
inp->inp_options = ip_srcroute();
tp = intotcpcb(inp);
tp->t_state = TCPS_LISTEN;
tp->t_flags |= tp0->t_flags & (TF_NOPUSH|TF_NOOPT);
/* Compute proper scaling value from buffer space */
while (tp->request_r_scale < TCP_MAX_WINSHIFT &&
TCP_MAXWIN << tp->request_r_scale <
so->so_rcv.sb_hiwat)
tp->request_r_scale++;
}
}
/*
* Segment received on connection.
* Reset idle time and keep-alive timer.
*/
tp->t_rcvtime = ticks;
if (TCPS_HAVEESTABLISHED(tp->t_state))
callout_reset(tp->tt_keep, tcp_keepidle, tcp_timer_keep, tp);
/*
* Process options if not in LISTEN state,
* else do it below (after getting remote address).
*/
if (tp->t_state != TCPS_LISTEN)
tcp_dooptions(tp, optp, optlen, ti, &to);
/*
* Header prediction: check for the two common cases
* of a uni-directional data xfer. If the packet has
* no control flags, is in-sequence, the window didn't
* change and we're not retransmitting, it's a
* candidate. If the length is zero and the ack moved
* forward, we're the sender side of the xfer. Just
* free the data acked & wake any higher level process
* that was blocked waiting for space. If the length
* is non-zero and the ack didn't move, we're the
* receiver side. If we're getting packets in-order
* (the reassembly queue is empty), add the data to
* the socket buffer and note that we need a delayed ack.
* Make sure that the hidden state-flags are also off.
* Since we check for TCPS_ESTABLISHED above, it can only
* be TH_NEEDSYN.
*/
if (tp->t_state == TCPS_ESTABLISHED &&
(tiflags & (TH_SYN|TH_FIN|TH_RST|TH_URG|TH_ACK)) == TH_ACK &&
((tp->t_flags & (TF_NEEDSYN|TF_NEEDFIN)) == 0) &&
((to.to_flag & TOF_TS) == 0 ||
TSTMP_GEQ(to.to_tsval, tp->ts_recent)) &&
/*
* Using the CC option is compulsory if once started:
* the segment is OK if no T/TCP was negotiated or
* if the segment has a CC option equal to CCrecv
*/
((tp->t_flags & (TF_REQ_CC|TF_RCVD_CC)) != (TF_REQ_CC|TF_RCVD_CC) ||
((to.to_flag & TOF_CC) != 0 && to.to_cc == tp->cc_recv)) &&
ti->ti_seq == tp->rcv_nxt &&
tiwin && tiwin == tp->snd_wnd &&
tp->snd_nxt == tp->snd_max) {
/*
* If last ACK falls within this segment's sequence numbers,
* record the timestamp.
* NOTE that the test is modified according to the latest
* proposal of the tcplw@cray.com list (Braden 1993/04/26).
*/
if ((to.to_flag & TOF_TS) != 0 &&
SEQ_LEQ(ti->ti_seq, tp->last_ack_sent)) {
tp->ts_recent_age = ticks;
tp->ts_recent = to.to_tsval;
}
if (ti->ti_len == 0) {
if (SEQ_GT(ti->ti_ack, tp->snd_una) &&
SEQ_LEQ(ti->ti_ack, tp->snd_max) &&
tp->snd_cwnd >= tp->snd_wnd &&
tp->t_dupacks < tcprexmtthresh) {
/*
* this is a pure ack for outstanding data.
*/
++tcpstat.tcps_predack;
/*
* "bad retransmit" recovery
*/
if (tp->t_rxtshift == 1 &&
ticks < tp->t_badrxtwin) {
tp->snd_cwnd = tp->snd_cwnd_prev;
tp->snd_ssthresh =
tp->snd_ssthresh_prev;
tp->snd_nxt = tp->snd_max;
tp->t_badrxtwin = 0;
}
if ((to.to_flag & TOF_TS) != 0)
tcp_xmit_timer(tp,
ticks - to.to_tsecr + 1);
else if (tp->t_rtttime &&
SEQ_GT(ti->ti_ack, tp->t_rtseq))
tcp_xmit_timer(tp, ticks - tp->t_rtttime);
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)
callout_stop(tp->tt_rexmt);
else if (!callout_active(tp->tt_persist))
callout_reset(tp->tt_rexmt,
tp->t_rxtcur,
tcp_timer_rexmt, tp);
sowwakeup(so);
if (so->so_snd.sb_cc)
(void) tcp_output(tp);
return;
}
} else if (ti->ti_ack == tp->snd_una &&
tp->t_segq == NULL &&
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);
if (tcp_delack_enabled) {
callout_reset(tp->tt_delack, tcp_delacktime,
tcp_timer_delack, tp);
} else {
tp->t_flags |= TF_ACKNOW;
tcp_output(tp);
}
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 = imax(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.
* If it is from this socket, drop it, it must be forged.
* 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: {
register struct sockaddr_in *sin;
if (tiflags & TH_RST)
goto drop;
if (tiflags & TH_ACK)
goto dropwithreset;
if ((tiflags & TH_SYN) == 0)
goto drop;
if ((ti->ti_dport == ti->ti_sport) &&
(ti->ti_dst.s_addr == ti->ti_src.s_addr))
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;
MALLOC(sin, struct sockaddr_in *, sizeof *sin, M_SONAME,
M_NOWAIT);
if (sin == NULL)
goto drop;
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, (struct sockaddr *)sin, &proc0)) {
inp->inp_laddr = laddr;
FREE(sin, M_SONAME);
goto drop;
}
FREE(sin, M_SONAME);
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/4;
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 (((tp->t_flags & TF_NOPUSH) != 0) &&
taop->tao_cc != 0 && CC_GT(to.to_cc, taop->tao_cc)) {
taop->tao_cc = to.to_cc;
tp->t_starttime = ticks;
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 (tcp_delack_enabled && ((tiflags & TH_FIN) ||
(ti->ti_len != 0 &&
in_localaddr(inp->inp_faddr)))) {
callout_reset(tp->tt_delack, tcp_delacktime,
tcp_timer_delack, tp);
tp->t_flags |= 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);
callout_reset(tp->tt_keep, tcp_keepinit,
tcp_timer_keep, tp);
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;
callout_reset(tp->tt_keep, tcp_keepinit, tcp_timer_keep, tp);
dropsocket = 0; /* committed to socket */
tcpstat.tcps_accepts++;
goto trimthenstep6;
}
/*
* If the state is SYN_RECEIVED:
* if seg contains an ACK, but not for our SYN/ACK, send a RST.
*/
case TCPS_SYN_RECEIVED:
if ((tiflags & TH_ACK) &&
(SEQ_LEQ(ti->ti_ack, tp->snd_una) ||
SEQ_GT(ti->ti_ack, tp->snd_max)))
goto dropwithreset;
break;
/*
* 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_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 ((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 no CC.ECHO option, make sure
* we don't get fooled into using T/TCP.
*/
if (to.to_flag & TOF_CCECHO) {
if (tp->cc_send != to.to_ccecho) {
if (taop->tao_ccsent != 0)
goto drop;
else
goto dropwithreset;
}
} else
tp->t_flags &= ~TF_RCVD_CC;
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 (tcp_delack_enabled && ti->ti_len != 0)
callout_reset(tp->tt_delack, tcp_delacktime,
tcp_timer_delack, tp);
else
tp->t_flags |= TF_ACKNOW;
/*
* Received <SYN,ACK> in SYN_SENT[*] state.
* Transitions:
* SYN_SENT --> ESTABLISHED
* SYN_SENT* --> FIN_WAIT_1
*/
tp->t_starttime = ticks;
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;
callout_reset(tp->tt_keep, tcp_keepidle,
tcp_timer_keep, tp);
}
} 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;
callout_stop(tp->tt_rexmt);
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;
tp->t_starttime = ticks;
if (tp->t_flags & TF_NEEDFIN) {
tp->t_state = TCPS_FIN_WAIT_1;
tp->t_flags &= ~TF_NEEDFIN;
} else {
tp->t_state = TCPS_ESTABLISHED;
callout_reset(tp->tt_keep,
tcp_keepidle,
tcp_timer_keep,
tp);
}
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 &&
(ticks - tp->t_starttime) > tcp_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 the RST flag and sequence number since reset segments
* are exempt from the timestamp and connection count tests. This
* fixes a bug introduced by the Stevens, vol. 2, p. 960 bugfix
* below which allowed reset segments in half the sequence space
* to fall though and be processed (which gives forged reset
* segments with a random sequence number a 50 percent chance of
* killing a connection).
* Then 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.
*
*
* If the RST bit is set, check the sequence number to see
* if this is a valid reset segment.
* RFC 793 page 37:
* In all states except SYN-SENT, all reset (RST) segments
* are validated by checking their SEQ-fields. A reset is
* valid if its sequence number is in the window.
* Note: this does not take into account delayed ACKs, so
* we should test against last_ack_sent instead of rcv_nxt.
* Also, it does not make sense to allow reset segments with
* sequence numbers greater than last_ack_sent to be processed
* since these sequence numbers are just the acknowledgement
* numbers in our outgoing packets being echoed back at us,
* and these acknowledgement numbers are monotonically
* increasing.
* If we have multiple segments in flight, the intial reset
* segment sequence numbers will be to the left of last_ack_sent,
* but they will eventually catch up.
* In any case, it never made sense to trim reset segments to
* fit the receive window since RFC 1122 says:
* 4.2.2.12 RST Segment: RFC-793 Section 3.4
*
* A TCP SHOULD allow a received RST segment to include data.
*
* DISCUSSION
* It has been suggested that a RST segment could contain
* ASCII text that encoded and explained the cause of the
* RST. No standard has yet been established for such
* data.
*
* If the reset segment passes the sequence number test examine
* the state:
* SYN_RECEIVED STATE:
* If passive open, return to LISTEN state.
* If active open, inform user that connection was refused.
* ESTABLISHED, FIN_WAIT_1, FIN_WAIT2, CLOSE_WAIT STATES:
* Inform user that connection was reset, and close tcb.
* CLOSING, LAST_ACK STATES:
* Close the tcb.
* TIME_WAIT STATE:
* Drop the segment - see Stevens, vol. 2, p. 964 and
* RFC 1337.
*/
if (tiflags & TH_RST) {
if (tp->last_ack_sent == ti->ti_seq) {
switch (tp->t_state) {
case TCPS_SYN_RECEIVED:
so->so_error = ECONNREFUSED;
goto close;
case TCPS_ESTABLISHED:
case TCPS_FIN_WAIT_1:
case TCPS_FIN_WAIT_2:
case TCPS_CLOSE_WAIT:
so->so_error = ECONNRESET;
close:
tp->t_state = TCPS_CLOSED;
tcpstat.tcps_drops++;
tp = tcp_close(tp);
break;
case TCPS_CLOSING:
case TCPS_LAST_ACK:
tp = tcp_close(tp);
break;
case TCPS_TIME_WAIT:
break;
}
}
goto drop;
}
/*
* 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 && tp->ts_recent &&
TSTMP_LT(to.to_tsval, tp->ts_recent)) {
/* Check to see if ts_recent is over 24 days old. */
if ((int)(ticks - 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 its 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))
goto dropafterack;
/*
* In the SYN-RECEIVED state, validate that the packet belongs to
* this connection before trimming the data to fit the receive
* window. Check the sequence number versus IRS since we know
* the sequence numbers haven't wrapped. This is a partial fix
* for the "LAND" DoS attack.
*/
if (tp->t_state == TCPS_SYN_RECEIVED && SEQ_LT(ti->ti_seq, tp->irs))
goto dropwithreset;
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.
*/
todrop = (ti->ti_seq+ti->ti_len) - (tp->rcv_nxt+tp->rcv_wnd);
if (todrop > 0) {
tcpstat.tcps_rcvpackafterwin++;
if (todrop >= ti->ti_len) {
tcpstat.tcps_rcvbyteafterwin += ti->ti_len;
/*
* If a new connection request is received
* while in TIME_WAIT, drop the old connection
* and start over if the sequence numbers
* are above the previous ones.
*/
if (tiflags & TH_SYN &&
tp->t_state == TCPS_TIME_WAIT &&
SEQ_GT(ti->ti_seq, tp->rcv_nxt)) {
iss = tp->snd_nxt + TCP_ISSINCR;
tp = tcp_close(tp);
goto findpcb;
}
/*
* If window is closed can only take segments at
* window edge, and have to drop data and PUSH from
* incoming segments. Continue processing, but
* remember to ack. Otherwise, drop segment
* and ack.
*/
if (tp->rcv_wnd == 0 && ti->ti_seq == tp->rcv_nxt) {
tp->t_flags |= TF_ACKNOW;
tcpstat.tcps_rcvwinprobe++;
} else
goto dropafterack;
} else
tcpstat.tcps_rcvbyteafterwin += todrop;
m_adj(m, -todrop);
ti->ti_len -= todrop;
tiflags &= ~(TH_PUSH|TH_FIN);
}
/*
* If last ACK falls within this segment's sequence numbers,
* record its timestamp.
* NOTE that the test is modified according to the latest
* proposal of the tcplw@cray.com list (Braden 1993/04/26).
*/
if ((to.to_flag & TOF_TS) != 0 &&
SEQ_LEQ(ti->ti_seq, tp->last_ack_sent)) {
tp->ts_recent_age = ticks;
tp->ts_recent = to.to_tsval;
}
/*
* If a SYN is in the window, then this is an
* error and we send an RST and drop the connection.
*/
if (tiflags & TH_SYN) {
tp = tcp_drop(tp, ECONNRESET);
goto dropwithreset;
}
/*
* If the ACK bit is off: if in SYN-RECEIVED state or SENDSYN
* flag is on (half-synchronized state), then queue data for
* later processing; else drop segment and return.
*/
if ((tiflags & TH_ACK) == 0) {
if (tp->t_state == TCPS_SYN_RECEIVED ||
(tp->t_flags & TF_NEEDSYN))
goto step6;
else
goto drop;
}
/*
* Ack processing.
*/
switch (tp->t_state) {
/*
* In SYN_RECEIVED state, the ack ACKs our SYN, so enter
* ESTABLISHED state and continue processing.
* The ACK was checked above.
*/
case TCPS_SYN_RECEIVED:
tcpstat.tcps_connects++;
soisconnected(so);
/* Do window scaling? */
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;
}
/*
* Upon successful completion of 3-way handshake,
* update cache.CC if it was undefined, pass any queued
* data to the user, and advance state appropriately.
*/
if ((taop = tcp_gettaocache(inp)) != NULL &&
taop->tao_cc == 0)
taop->tao_cc = tp->cc_recv;
/*
* Make transitions:
* SYN-RECEIVED -> ESTABLISHED
* SYN-RECEIVED* -> FIN-WAIT-1
*/
tp->t_starttime = ticks;
if (tp->t_flags & TF_NEEDFIN) {
tp->t_state = TCPS_FIN_WAIT_1;
tp->t_flags &= ~TF_NEEDFIN;
} else {
tp->t_state = TCPS_ESTABLISHED;
callout_reset(tp->tt_keep, tcp_keepidle,
tcp_timer_keep, tp);
}
/*
* If segment contains data or ACK, will call tcp_reass()
* later; if not, do so now to pass queued data to user.
*/
if (ti->ti_len == 0 && (tiflags & TH_FIN) == 0)
(void) tcp_reass(tp, (struct tcpiphdr *)0,
(struct mbuf *)0);
tp->snd_wl1 = ti->ti_seq - 1;
/* fall into ... */
/*
* In ESTABLISHED state: drop duplicate ACKs; ACK out of range
* ACKs. If the ack is in the range
* tp->snd_una < ti->ti_ack <= tp->snd_max
* then advance tp->snd_una to ti->ti_ack and drop
* data from the retransmission queue. If this ACK reflects
* more up to date window information we update our window information.
*/
case TCPS_ESTABLISHED:
case TCPS_FIN_WAIT_1:
case TCPS_FIN_WAIT_2:
case TCPS_CLOSE_WAIT:
case TCPS_CLOSING:
case TCPS_LAST_ACK:
case TCPS_TIME_WAIT:
if (SEQ_LEQ(ti->ti_ack, tp->snd_una)) {
if (ti->ti_len == 0 && tiwin == tp->snd_wnd) {
tcpstat.tcps_rcvdupack++;
/*
* If we have outstanding data (other than
* a window probe), this is a completely
* duplicate ack (ie, window info didn't
* change), the ack is the biggest we've
* seen and we've seen exactly our rexmt
* threshhold of them, assume a packet
* has been dropped and retransmit it.
* Kludge snd_nxt & the congestion
* window so we send only this one
* packet.
*
* We know we're losing at the current
* window size so do congestion avoidance
* (set ssthresh to half the current window
* and pull our congestion window back to
* the new ssthresh).
*
* Dup acks mean that packets have left the
* network (they're now cached at the receiver)
* so bump cwnd by the amount in the receiver
* to keep a constant cwnd packets in the
* network.
*/
if (!callout_active(tp->tt_rexmt) ||
ti->ti_ack != tp->snd_una)
tp->t_dupacks = 0;
else if (++tp->t_dupacks == tcprexmtthresh) {
tcp_seq onxt = tp->snd_nxt;
u_int win =
min(tp->snd_wnd, tp->snd_cwnd) / 2 /
tp->t_maxseg;
if (win < 2)
win = 2;
tp->snd_ssthresh = win * tp->t_maxseg;
callout_stop(tp->tt_rexmt);
tp->t_rtttime = 0;
tp->snd_nxt = ti->ti_ack;
tp->snd_cwnd = tp->t_maxseg;
(void) tcp_output(tp);
tp->snd_cwnd = tp->snd_ssthresh +
tp->t_maxseg * tp->t_dupacks;
if (SEQ_GT(onxt, tp->snd_nxt))
tp->snd_nxt = onxt;
goto drop;
} else if (tp->t_dupacks > tcprexmtthresh) {
tp->snd_cwnd += tp->t_maxseg;
(void) tcp_output(tp);
goto drop;
}
} else
tp->t_dupacks = 0;
break;
}
/*
* If the congestion window was inflated to account
* for the other side's cached packets, retract it.
*/
if (tp->t_dupacks >= tcprexmtthresh &&
tp->snd_cwnd > tp->snd_ssthresh)
tp->snd_cwnd = tp->snd_ssthresh;
tp->t_dupacks = 0;
if (SEQ_GT(ti->ti_ack, tp->snd_max)) {
tcpstat.tcps_rcvacktoomuch++;
goto dropafterack;
}
/*
* If we reach this point, ACK is not a duplicate,
* i.e., it ACKs something we sent.
*/
if (tp->t_flags & TF_NEEDSYN) {
/*
* T/TCP: Connection was half-synchronized, and our
* SYN has been ACK'd (so connection is now fully
* synchronized). Go to non-starred state,
* increment snd_una for ACK of SYN, and check if
* we can do window scaling.
*/
tp->t_flags &= ~TF_NEEDSYN;
tp->snd_una++;
/* Do window scaling? */
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;
}
}
process_ACK:
acked = ti->ti_ack - tp->snd_una;
tcpstat.tcps_rcvackpack++;
tcpstat.tcps_rcvackbyte += acked;
/*
* If we just performed our first retransmit, and the ACK
* arrives within our recovery window, then it was a mistake
* to do the retransmit in the first place. Recover our
* original cwnd and ssthresh, and proceed to transmit where
* we left off.
*/
if (tp->t_rxtshift == 1 && ticks < tp->t_badrxtwin) {
tp->snd_cwnd = tp->snd_cwnd_prev;
tp->snd_ssthresh = tp->snd_ssthresh_prev;
tp->snd_nxt = tp->snd_max;
tp->t_badrxtwin = 0; /* XXX probably not required */
}
/*
* If we have a timestamp reply, update smoothed
* round trip time. If no timestamp is present but
* transmit timer is running and timed sequence
* number was acked, update smoothed round trip time.
* Since we now have an rtt measurement, cancel the
* timer backoff (cf., Phil Karn's retransmit alg.).
* Recompute the initial retransmit timer.
*/
if (to.to_flag & TOF_TS)
tcp_xmit_timer(tp, ticks - to.to_tsecr + 1);
else if (tp->t_rtttime && SEQ_GT(ti->ti_ack, tp->t_rtseq))
tcp_xmit_timer(tp, ticks - tp->t_rtttime);
/*
* If all outstanding data is acked, stop retransmit
* timer and remember to restart (more output or persist).
* If there is more data to be acked, restart retransmit
* timer, using current (possibly backed-off) value.
*/
if (ti->ti_ack == tp->snd_max) {
callout_stop(tp->tt_rexmt);
needoutput = 1;
} else if (!callout_active(tp->tt_persist))
callout_reset(tp->tt_rexmt, tp->t_rxtcur,
tcp_timer_rexmt, tp);
/*
* If no data (only SYN) was ACK'd,
* skip rest of ACK processing.
*/
if (acked == 0)
goto step6;
/*
* When new data is acked, open the congestion window.
* If the window gives us less than ssthresh packets
* in flight, open exponentially (maxseg per packet).
* Otherwise open linearly: maxseg per window
* (maxseg^2 / cwnd per packet).
*/
{
register u_int cw = tp->snd_cwnd;
register u_int incr = tp->t_maxseg;
if (cw > tp->snd_ssthresh)
incr = incr * incr / cw;
tp->snd_cwnd = min(cw + incr, TCP_MAXWIN << tp->snd_scale);
}
if (acked > so->so_snd.sb_cc) {
tp->snd_wnd -= so->so_snd.sb_cc;
sbdrop(&so->so_snd, (int)so->so_snd.sb_cc);
ourfinisacked = 1;
} else {
sbdrop(&so->so_snd, acked);
tp->snd_wnd -= acked;
ourfinisacked = 0;
}
sowwakeup(so);
tp->snd_una = ti->ti_ack;
if (SEQ_LT(tp->snd_nxt, tp->snd_una))
tp->snd_nxt = tp->snd_una;
switch (tp->t_state) {
/*
* In FIN_WAIT_1 STATE in addition to the processing
* for the ESTABLISHED state if our FIN is now acknowledged
* then enter FIN_WAIT_2.
*/
case TCPS_FIN_WAIT_1:
if (ourfinisacked) {
/*
* If we can't receive any more
* data, then closing user can proceed.
* Starting the timer is contrary to the
* specification, but if we don't get a FIN
* we'll hang forever.
*/
if (so->so_state & SS_CANTRCVMORE) {
soisdisconnected(so);
callout_reset(tp->tt_2msl, tcp_maxidle,
tcp_timer_2msl, tp);
}
tp->t_state = TCPS_FIN_WAIT_2;
}
break;
/*
* In CLOSING STATE in addition to the processing for
* the ESTABLISHED state if the ACK acknowledges our FIN
* then enter the TIME-WAIT state, otherwise ignore
* the segment.
*/
case TCPS_CLOSING:
if (ourfinisacked) {
tp->t_state = TCPS_TIME_WAIT;
tcp_canceltimers(tp);
/* Shorten TIME_WAIT [RFC-1644, p.28] */
if (tp->cc_recv != 0 &&
(ticks - tp->t_starttime) < tcp_msl)
callout_reset(tp->tt_2msl,
tp->t_rxtcur *
TCPTV_TWTRUNC,
tcp_timer_2msl, tp);
else
callout_reset(tp->tt_2msl, 2 * tcp_msl,
tcp_timer_2msl, tp);
soisdisconnected(so);
}
break;
/*
* In LAST_ACK, we may still be waiting for data to drain
* and/or to be acked, as well as for the ack of our FIN.
* If our FIN is now acknowledged, delete the TCB,
* enter the closed state and return.
*/
case TCPS_LAST_ACK:
if (ourfinisacked) {
tp = tcp_close(tp);
goto drop;
}
break;
/*
* In TIME_WAIT state the only thing that should arrive
* is a retransmission of the remote FIN. Acknowledge
* it and restart the finack timer.
*/
case TCPS_TIME_WAIT:
callout_reset(tp->tt_2msl, 2 * tcp_msl,
tcp_timer_2msl, tp);
goto dropafterack;
}
}
step6:
/*
* Update window information.
* Don't look at window if no ACK: TAC's send garbage on first SYN.
*/
if ((tiflags & TH_ACK) &&
(SEQ_LT(tp->snd_wl1, ti->ti_seq) ||
(tp->snd_wl1 == ti->ti_seq && (SEQ_LT(tp->snd_wl2, ti->ti_ack) ||
(tp->snd_wl2 == ti->ti_ack && tiwin > tp->snd_wnd))))) {
/* keep track of pure window updates */
if (ti->ti_len == 0 &&
tp->snd_wl2 == ti->ti_ack && tiwin > tp->snd_wnd)
tcpstat.tcps_rcvwinupd++;
tp->snd_wnd = tiwin;
tp->snd_wl1 = ti->ti_seq;
tp->snd_wl2 = ti->ti_ack;
if (tp->snd_wnd > tp->max_sndwnd)
tp->max_sndwnd = tp->snd_wnd;
needoutput = 1;
}
/*
* Process segments with URG.
*/
if ((tiflags & TH_URG) && ti->ti_urp &&
TCPS_HAVERCVDFIN(tp->t_state) == 0) {
/*
* This is a kludge, but if we receive and accept
* random urgent pointers, we'll crash in
* soreceive. It's hard to imagine someone
* actually wanting to send this much urgent data.
*/
if (ti->ti_urp + so->so_rcv.sb_cc > sb_max) {
ti->ti_urp = 0; /* XXX */
tiflags &= ~TH_URG; /* XXX */
goto dodata; /* XXX */
}
/*
* If this segment advances the known urgent pointer,
* then mark the data stream. This should not happen
* in CLOSE_WAIT, CLOSING, LAST_ACK or TIME_WAIT STATES since
* a FIN has been received from the remote side.
* In these states we ignore the URG.
*
* According to RFC961 (Assigned Protocols),
* the urgent pointer points to the last octet
* of urgent data. We continue, however,
* to consider it to indicate the first octet
* of data past the urgent section as the original
* spec states (in one of two places).
*/
if (SEQ_GT(ti->ti_seq+ti->ti_urp, tp->rcv_up)) {
tp->rcv_up = ti->ti_seq + ti->ti_urp;
so->so_oobmark = so->so_rcv.sb_cc +
(tp->rcv_up - tp->rcv_nxt) - 1;
if (so->so_oobmark == 0)
so->so_state |= SS_RCVATMARK;
sohasoutofband(so);
tp->t_oobflags &= ~(TCPOOB_HAVEDATA | TCPOOB_HADDATA);
}
/*
* Remove out of band data so doesn't get presented to user.
* This can happen independent of advancing the URG pointer,
* but if two URG's are pending at once, some out-of-band
* data may creep in... ick.
*/
if (ti->ti_urp <= (u_long)ti->ti_len
#ifdef SO_OOBINLINE
&& (so->so_options & SO_OOBINLINE) == 0
#endif
)
tcp_pulloutofband(so, ti, m);
} else
/*
* If no out of band data is expected,
* pull receive urgent pointer along
* with the receive window.
*/
if (SEQ_GT(tp->rcv_nxt, tp->rcv_up))
tp->rcv_up = tp->rcv_nxt;
dodata: /* XXX */
/*
* Process the segment text, merging it into the TCP sequencing queue,
* and arranging for acknowledgment of receipt if necessary.
* This process logically involves adjusting tp->rcv_wnd as data
* is presented to the user (this happens in tcp_usrreq.c,
* case PRU_RCVD). If a FIN has already been received on this
* connection then we just ignore the text.
*/
if ((ti->ti_len || (tiflags&TH_FIN)) &&
TCPS_HAVERCVDFIN(tp->t_state) == 0) {
TCP_REASS(tp, ti, m, so, tiflags);
/*
* Note the amount of data that peer has sent into
* our window, in order to estimate the sender's
* buffer size.
*/
len = so->so_rcv.sb_hiwat - (tp->rcv_adv - tp->rcv_nxt);
} else {
m_freem(m);
tiflags &= ~TH_FIN;
}
/*
* If FIN is received ACK the FIN and let the user know
* that the connection is closing.
*/
if (tiflags & TH_FIN) {
if (TCPS_HAVERCVDFIN(tp->t_state) == 0) {
socantrcvmore(so);
/*
* If connection is half-synchronized
* (ie NEEDSYN flag on) then delay ACK,
* so it may be piggybacked when SYN is sent.
* Otherwise, since we received a FIN then no
* more input can be expected, send ACK now.
*/
if (tcp_delack_enabled && (tp->t_flags & TF_NEEDSYN))
callout_reset(tp->tt_delack, tcp_delacktime,
tcp_timer_delack, tp);
else
tp->t_flags |= TF_ACKNOW;
tp->rcv_nxt++;
}
switch (tp->t_state) {
/*
* In SYN_RECEIVED and ESTABLISHED STATES
* enter the CLOSE_WAIT state.
*/
case TCPS_SYN_RECEIVED:
tp->t_starttime = ticks;
/*FALLTHROUGH*/
case TCPS_ESTABLISHED:
tp->t_state = TCPS_CLOSE_WAIT;
break;
/*
* If still in FIN_WAIT_1 STATE FIN has not been acked so
* enter the CLOSING state.
*/
case TCPS_FIN_WAIT_1:
tp->t_state = TCPS_CLOSING;
break;
/*
* In FIN_WAIT_2 state enter the TIME_WAIT state,
* starting the time-wait timer, turning off the other
* standard timers.
*/
case TCPS_FIN_WAIT_2:
tp->t_state = TCPS_TIME_WAIT;
tcp_canceltimers(tp);
/* Shorten TIME_WAIT [RFC-1644, p.28] */
if (tp->cc_recv != 0 &&
(ticks - tp->t_starttime) < tcp_msl) {
callout_reset(tp->tt_2msl,
tp->t_rxtcur * TCPTV_TWTRUNC,
tcp_timer_2msl, tp);
/* For transaction client, force ACK now. */
tp->t_flags |= TF_ACKNOW;
}
else
callout_reset(tp->tt_2msl, 2 * tcp_msl,
tcp_timer_2msl, tp);
soisdisconnected(so);
break;
/*
* In TIME_WAIT state restart the 2 MSL time_wait timer.
*/
case TCPS_TIME_WAIT:
callout_reset(tp->tt_2msl, 2 * tcp_msl,
tcp_timer_2msl, tp);
break;
}
}
#ifdef TCPDEBUG
if (so->so_options & SO_DEBUG)
tcp_trace(TA_INPUT, ostate, tp, &tcp_saveti, 0);
#endif
/*
* Return any desired output.
*/
if (needoutput || (tp->t_flags & TF_ACKNOW))
(void) tcp_output(tp);
return;
dropafterack:
/*
* Generate an ACK dropping incoming segment if it occupies
* sequence space, where the ACK reflects our state.
*
* We can now skip the test for the RST flag since all
* paths to this code happen after packets containing
* RST have been dropped.
*
* In the SYN-RECEIVED state, don't send an ACK unless the
* segment we received passes the SYN-RECEIVED ACK test.
* If it fails send a RST. This breaks the loop in the
* "LAND" DoS attack, and also prevents an ACK storm
* between two listening ports that have been sent forged
* SYN segments, each with the source address of the other.
*/
if (tp->t_state == TCPS_SYN_RECEIVED && (tiflags & TH_ACK) &&
(SEQ_GT(tp->snd_una, ti->ti_ack) ||
SEQ_GT(ti->ti_ack, tp->snd_max)) )
goto dropwithreset;
#ifdef TCPDEBUG
if (so->so_options & SO_DEBUG)
tcp_trace(TA_DROP, ostate, tp, &tcp_saveti, 0);
#endif
m_freem(m);
tp->t_flags |= TF_ACKNOW;
(void) tcp_output(tp);
return;
dropwithreset:
#ifdef TCP_RESTRICT_RST
if (restrict_rst)
goto drop;
#endif
/*
* Generate a RST, dropping incoming segment.
* Make ACK acceptable to originator of segment.
* Don't bother to respond if destination was broadcast/multicast.
*/
if ((tiflags & TH_RST) || m->m_flags & (M_BCAST|M_MCAST) ||
IN_MULTICAST(ntohl(ti->ti_dst.s_addr)))
goto drop;
#ifdef TCPDEBUG
if (tp == 0 || (tp->t_inpcb->inp_socket->so_options & SO_DEBUG))
tcp_trace(TA_DROP, ostate, tp, &tcp_saveti, 0);
#endif
if (tiflags & TH_ACK)
tcp_respond(tp, ti, m, (tcp_seq)0, ti->ti_ack, TH_RST);
else {
if (tiflags & TH_SYN)
ti->ti_len++;
tcp_respond(tp, ti, m, ti->ti_seq+ti->ti_len, (tcp_seq)0,
TH_RST|TH_ACK);
}
/* destroy temporarily created socket */
if (dropsocket)
(void) soabort(so);
return;
drop:
/*
* Drop space held by incoming segment and return.
*/
#ifdef TCPDEBUG
if (tp == 0 || (tp->t_inpcb->inp_socket->so_options & SO_DEBUG))
tcp_trace(TA_DROP, ostate, tp, &tcp_saveti, 0);
#endif
m_freem(m);
/* destroy temporarily created socket */
if (dropsocket)
(void) soabort(so);
return;
}
static void
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 = ticks;
}
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.
*/
static 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.
*/
static void
tcp_xmit_timer(tp, rtt)
register struct tcpcb *tp;
int rtt;
{
register int delta;
tcpstat.tcps_rttupdated++;
tp->t_rttupdated++;
if (tp->t_srtt != 0) {
/*
* srtt is stored as fixed point with 5 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) << TCP_DELTA_SHIFT)
- (tp->t_srtt >> (TCP_RTT_SHIFT - TCP_DELTA_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 4 bits after the
* binary point (scaled by 16). 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 - TCP_DELTA_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_rtttime = 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),
max(tp->t_rttmin, rtt + 2), 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;
}
ifp = rt->rt_ifp;
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 / hz);
tp->t_srtt = rtt / (RTM_RTTUNIT / (hz * TCP_RTT_SCALE));
tcpstat.tcps_usedrtt++;
if (rt->rt_rmx.rmx_rttvar) {
tp->t_rttvar = rt->rt_rmx.rmx_rttvar /
(RTM_RTTUNIT / (hz * 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, so, NULL);
}
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, so, NULL);
}
/*
* Set the slow-start flight size depending on whether this
* is a local network or not.
*/
if (in_localaddr(inp->inp_faddr))
tp->snd_cwnd = mss * ss_fltsz_local;
else
tp->snd_cwnd = mss * ss_fltsz;
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;
return rt->rt_ifp->if_mtu - sizeof(struct tcpiphdr);
}
Reference in New Issue View Git Blame Copy Permalink