tcp_var.h revision 2169
1/*
2 * Copyright (c) 1982, 1986, 1993, 1994
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 *    notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 *    notice, this list of conditions and the following disclaimer in the
12 *    documentation and/or other materials provided with the distribution.
13 * 3. All advertising materials mentioning features or use of this software
14 *    must display the following acknowledgement:
15 *	This product includes software developed by the University of
16 *	California, Berkeley and its contributors.
17 * 4. Neither the name of the University nor the names of its contributors
18 *    may be used to endorse or promote products derived from this software
19 *    without specific prior written permission.
20 *
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 * SUCH DAMAGE.
32 *
33 *	@(#)tcp_var.h	8.3 (Berkeley) 4/10/94
34 * $Id: tcp_var.h,v 1.2 1994/08/02 07:49:17 davidg Exp $
35 */
36
37#ifndef _NETINET_TCP_VAR_H_
38#define _NETINET_TCP_VAR_H_
39/*
40 * Kernel variables for tcp.
41 */
42
43/*
44 * Tcp control block, one per tcp; fields:
45 */
46struct tcpcb {
47	struct	tcpiphdr *seg_next;	/* sequencing queue */
48	struct	tcpiphdr *seg_prev;
49	short	t_state;		/* state of this connection */
50	short	t_timer[TCPT_NTIMERS];	/* tcp timers */
51	short	t_rxtshift;		/* log(2) of rexmt exp. backoff */
52	short	t_rxtcur;		/* current retransmit value */
53	short	t_dupacks;		/* consecutive dup acks recd */
54	u_short	t_maxseg;		/* maximum segment size */
55	char	t_force;		/* 1 if forcing out a byte */
56	u_short	t_flags;
57#define	TF_ACKNOW	0x0001		/* ack peer immediately */
58#define	TF_DELACK	0x0002		/* ack, but try to delay it */
59#define	TF_NODELAY	0x0004		/* don't delay packets to coalesce */
60#define	TF_NOOPT	0x0008		/* don't use tcp options */
61#define	TF_SENTFIN	0x0010		/* have sent FIN */
62#define	TF_REQ_SCALE	0x0020		/* have/will request window scaling */
63#define	TF_RCVD_SCALE	0x0040		/* other side has requested scaling */
64#define	TF_REQ_TSTMP	0x0080		/* have/will request timestamps */
65#define	TF_RCVD_TSTMP	0x0100		/* a timestamp was received in SYN */
66#define	TF_SACK_PERMIT	0x0200		/* other side said I could SACK */
67
68	struct	tcpiphdr *t_template;	/* skeletal packet for transmit */
69	struct	inpcb *t_inpcb;		/* back pointer to internet pcb */
70/*
71 * The following fields are used as in the protocol specification.
72 * See RFC783, Dec. 1981, page 21.
73 */
74/* send sequence variables */
75	tcp_seq	snd_una;		/* send unacknowledged */
76	tcp_seq	snd_nxt;		/* send next */
77	tcp_seq	snd_up;			/* send urgent pointer */
78	tcp_seq	snd_wl1;		/* window update seg seq number */
79	tcp_seq	snd_wl2;		/* window update seg ack number */
80	tcp_seq	iss;			/* initial send sequence number */
81	u_long	snd_wnd;		/* send window */
82/* receive sequence variables */
83	u_long	rcv_wnd;		/* receive window */
84	tcp_seq	rcv_nxt;		/* receive next */
85	tcp_seq	rcv_up;			/* receive urgent pointer */
86	tcp_seq	irs;			/* initial receive sequence number */
87/*
88 * Additional variables for this implementation.
89 */
90/* receive variables */
91	tcp_seq	rcv_adv;		/* advertised window */
92/* retransmit variables */
93	tcp_seq	snd_max;		/* highest sequence number sent;
94					 * used to recognize retransmits
95					 */
96/* congestion control (for slow start, source quench, retransmit after loss) */
97	u_long	snd_cwnd;		/* congestion-controlled window */
98	u_long	snd_ssthresh;		/* snd_cwnd size threshhold for
99					 * for slow start exponential to
100					 * linear switch
101					 */
102/*
103 * transmit timing stuff.  See below for scale of srtt and rttvar.
104 * "Variance" is actually smoothed difference.
105 */
106	short	t_idle;			/* inactivity time */
107	short	t_rtt;			/* round trip time */
108	tcp_seq	t_rtseq;		/* sequence number being timed */
109	short	t_srtt;			/* smoothed round-trip time */
110	short	t_rttvar;		/* variance in round-trip time */
111	u_short	t_rttmin;		/* minimum rtt allowed */
112	u_long	max_sndwnd;		/* largest window peer has offered */
113
114/* out-of-band data */
115	char	t_oobflags;		/* have some */
116	char	t_iobc;			/* input character */
117#define	TCPOOB_HAVEDATA	0x01
118#define	TCPOOB_HADDATA	0x02
119	short	t_softerror;		/* possible error not yet reported */
120
121/* RFC 1323 variables */
122	u_char	snd_scale;		/* window scaling for send window */
123	u_char	rcv_scale;		/* window scaling for recv window */
124	u_char	request_r_scale;	/* pending window scaling */
125	u_char	requested_s_scale;
126	u_long	ts_recent;		/* timestamp echo data */
127	u_long	ts_recent_age;		/* when last updated */
128	tcp_seq	last_ack_sent;
129
130/* TUBA stuff */
131	caddr_t	t_tuba_pcb;		/* next level down pcb for TCP over z */
132};
133
134#define	intotcpcb(ip)	((struct tcpcb *)(ip)->inp_ppcb)
135#define	sototcpcb(so)	(intotcpcb(sotoinpcb(so)))
136
137/*
138 * The smoothed round-trip time and estimated variance
139 * are stored as fixed point numbers scaled by the values below.
140 * For convenience, these scales are also used in smoothing the average
141 * (smoothed = (1/scale)sample + ((scale-1)/scale)smoothed).
142 * With these scales, srtt has 3 bits to the right of the binary point,
143 * and thus an "ALPHA" of 0.875.  rttvar has 2 bits to the right of the
144 * binary point, and is smoothed with an ALPHA of 0.75.
145 */
146#define	TCP_RTT_SCALE		8	/* multiplier for srtt; 3 bits frac. */
147#define	TCP_RTT_SHIFT		3	/* shift for srtt; 3 bits frac. */
148#define	TCP_RTTVAR_SCALE	4	/* multiplier for rttvar; 2 bits */
149#define	TCP_RTTVAR_SHIFT	2	/* multiplier for rttvar; 2 bits */
150
151/*
152 * The initial retransmission should happen at rtt + 4 * rttvar.
153 * Because of the way we do the smoothing, srtt and rttvar
154 * will each average +1/2 tick of bias.  When we compute
155 * the retransmit timer, we want 1/2 tick of rounding and
156 * 1 extra tick because of +-1/2 tick uncertainty in the
157 * firing of the timer.  The bias will give us exactly the
158 * 1.5 tick we need.  But, because the bias is
159 * statistical, we have to test that we don't drop below
160 * the minimum feasible timer (which is 2 ticks).
161 * This macro assumes that the value of TCP_RTTVAR_SCALE
162 * is the same as the multiplier for rttvar.
163 */
164#define	TCP_REXMTVAL(tp) \
165	(((tp)->t_srtt >> TCP_RTT_SHIFT) + (tp)->t_rttvar)
166
167/* XXX
168 * We want to avoid doing m_pullup on incoming packets but that
169 * means avoiding dtom on the tcp reassembly code.  That in turn means
170 * keeping an mbuf pointer in the reassembly queue (since we might
171 * have a cluster).  As a quick hack, the source & destination
172 * port numbers (which are no longer needed once we've located the
173 * tcpcb) are overlayed with an mbuf pointer.
174 */
175#define REASS_MBUF(ti) (*(struct mbuf **)&((ti)->ti_t))
176
177/*
178 * TCP statistics.
179 * Many of these should be kept per connection,
180 * but that's inconvenient at the moment.
181 */
182struct	tcpstat {
183	u_long	tcps_connattempt;	/* connections initiated */
184	u_long	tcps_accepts;		/* connections accepted */
185	u_long	tcps_connects;		/* connections established */
186	u_long	tcps_drops;		/* connections dropped */
187	u_long	tcps_conndrops;		/* embryonic connections dropped */
188	u_long	tcps_closed;		/* conn. closed (includes drops) */
189	u_long	tcps_segstimed;		/* segs where we tried to get rtt */
190	u_long	tcps_rttupdated;	/* times we succeeded */
191	u_long	tcps_delack;		/* delayed acks sent */
192	u_long	tcps_timeoutdrop;	/* conn. dropped in rxmt timeout */
193	u_long	tcps_rexmttimeo;	/* retransmit timeouts */
194	u_long	tcps_persisttimeo;	/* persist timeouts */
195	u_long	tcps_keeptimeo;		/* keepalive timeouts */
196	u_long	tcps_keepprobe;		/* keepalive probes sent */
197	u_long	tcps_keepdrops;		/* connections dropped in keepalive */
198
199	u_long	tcps_sndtotal;		/* total packets sent */
200	u_long	tcps_sndpack;		/* data packets sent */
201	u_long	tcps_sndbyte;		/* data bytes sent */
202	u_long	tcps_sndrexmitpack;	/* data packets retransmitted */
203	u_long	tcps_sndrexmitbyte;	/* data bytes retransmitted */
204	u_long	tcps_sndacks;		/* ack-only packets sent */
205	u_long	tcps_sndprobe;		/* window probes sent */
206	u_long	tcps_sndurg;		/* packets sent with URG only */
207	u_long	tcps_sndwinup;		/* window update-only packets sent */
208	u_long	tcps_sndctrl;		/* control (SYN|FIN|RST) packets sent */
209
210	u_long	tcps_rcvtotal;		/* total packets received */
211	u_long	tcps_rcvpack;		/* packets received in sequence */
212	u_long	tcps_rcvbyte;		/* bytes received in sequence */
213	u_long	tcps_rcvbadsum;		/* packets received with ccksum errs */
214	u_long	tcps_rcvbadoff;		/* packets received with bad offset */
215	u_long	tcps_rcvshort;		/* packets received too short */
216	u_long	tcps_rcvduppack;	/* duplicate-only packets received */
217	u_long	tcps_rcvdupbyte;	/* duplicate-only bytes received */
218	u_long	tcps_rcvpartduppack;	/* packets with some duplicate data */
219	u_long	tcps_rcvpartdupbyte;	/* dup. bytes in part-dup. packets */
220	u_long	tcps_rcvoopack;		/* out-of-order packets received */
221	u_long	tcps_rcvoobyte;		/* out-of-order bytes received */
222	u_long	tcps_rcvpackafterwin;	/* packets with data after window */
223	u_long	tcps_rcvbyteafterwin;	/* bytes rcvd after window */
224	u_long	tcps_rcvafterclose;	/* packets rcvd after "close" */
225	u_long	tcps_rcvwinprobe;	/* rcvd window probe packets */
226	u_long	tcps_rcvdupack;		/* rcvd duplicate acks */
227	u_long	tcps_rcvacktoomuch;	/* rcvd acks for unsent data */
228	u_long	tcps_rcvackpack;	/* rcvd ack packets */
229	u_long	tcps_rcvackbyte;	/* bytes acked by rcvd acks */
230	u_long	tcps_rcvwinupd;		/* rcvd window update packets */
231	u_long	tcps_pawsdrop;		/* segments dropped due to PAWS */
232	u_long	tcps_predack;		/* times hdr predict ok for acks */
233	u_long	tcps_preddat;		/* times hdr predict ok for data pkts */
234	u_long	tcps_pcbcachemiss;
235};
236
237#ifdef KERNEL
238struct	inpcb tcb;		/* head of queue of active tcpcb's */
239struct	tcpstat tcpstat;	/* tcp statistics */
240u_long	tcp_now;		/* for RFC 1323 timestamps */
241
242int	 tcp_attach __P((struct socket *));
243void	 tcp_canceltimers __P((struct tcpcb *));
244struct tcpcb *
245	 tcp_close __P((struct tcpcb *));
246void	 tcp_ctlinput __P((int, struct sockaddr *, struct ip *));
247int	 tcp_ctloutput __P((int, struct socket *, int, int, struct mbuf **));
248struct tcpcb *
249	 tcp_disconnect __P((struct tcpcb *));
250struct tcpcb *
251	 tcp_drop __P((struct tcpcb *, int));
252void	 tcp_dooptions __P((struct tcpcb *,
253	    u_char *, int, struct tcpiphdr *, int *, u_long *, u_long *));
254void	 tcp_drain __P((void));
255void	 tcp_fasttimo __P((void));
256void	 tcp_init __P((void));
257void	 tcp_input __P((struct mbuf *, int));
258int	 tcp_mss __P((struct tcpcb *, u_int));
259struct tcpcb *
260	 tcp_newtcpcb __P((struct inpcb *));
261void	 tcp_notify __P((struct inpcb *, int));
262int	 tcp_output __P((struct tcpcb *));
263void	 tcp_pulloutofband __P((struct socket *,
264	    struct tcpiphdr *, struct mbuf *));
265void	 tcp_quench __P((struct inpcb *, int));
266int	 tcp_reass __P((struct tcpcb *, struct tcpiphdr *, struct mbuf *));
267void	 tcp_respond __P((struct tcpcb *,
268	    struct tcpiphdr *, struct mbuf *, u_long, u_long, int));
269void	 tcp_setpersist __P((struct tcpcb *));
270void	 tcp_slowtimo __P((void));
271struct tcpiphdr *
272	 tcp_template __P((struct tcpcb *));
273struct tcpcb *
274	 tcp_timers __P((struct tcpcb *, int));
275void	 tcp_trace __P((int, int, struct tcpcb *, struct tcpiphdr *, int));
276struct tcpcb *
277	 tcp_usrclosed __P((struct tcpcb *));
278int	 tcp_usrreq __P((struct socket *,
279	    int, struct mbuf *, struct mbuf *, struct mbuf *));
280void	 tcp_xmit_timer __P((struct tcpcb *, int));
281#endif
282
283#endif
284