[unix-history] / sys / netinet / tcp_var.h

/*
 * Copyright (c) 1982, 1986 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.
 *
 *	from: @(#)tcp_var.h	7.10 (Berkeley) 6/28/90
 *	$Id$
 */

/*
 * Kernel variables for tcp.
 */

/*
 * Tcp control block, one per tcp; fields:
 */
struct tcpcb {
	struct	tcpiphdr *seg_next;	/* sequencing queue */
	struct	tcpiphdr *seg_prev;
	short	t_state;		/* state of this connection */
	short	t_timer[TCPT_NTIMERS];	/* tcp timers */
	short	t_rxtshift;		/* log(2) of rexmt exp. backoff */
	short	t_rxtcur;		/* current retransmit value */
	short	t_dupacks;		/* consecutive dup acks recd */
	u_short	t_maxseg;		/* maximum segment size */
	char	t_force;		/* 1 if forcing out a byte */
	u_char	t_flags;
#define	TF_ACKNOW	0x01		/* ack peer immediately */
#define	TF_DELACK	0x02		/* ack, but try to delay it */
#define	TF_NODELAY	0x04		/* don't delay packets to coalesce */
#define	TF_NOOPT	0x08		/* don't use tcp options */
#define	TF_SENTFIN	0x10		/* have sent FIN */
	struct	tcpiphdr *t_template;	/* skeletal packet for transmit */
	struct	inpcb *t_inpcb;		/* back pointer to internet pcb */
/*
 * The following fields are used as in the protocol specification.
 * See RFC783, Dec. 1981, page 21.
 */
/* send sequence variables */
	tcp_seq	snd_una;		/* send unacknowledged */
	tcp_seq	snd_nxt;		/* send next */
	tcp_seq	snd_up;			/* send urgent pointer */
	tcp_seq	snd_wl1;		/* window update seg seq number */
	tcp_seq	snd_wl2;		/* window update seg ack number */
	tcp_seq	iss;			/* initial send sequence number */
	u_short	snd_wnd;		/* send window */
/* receive sequence variables */
	u_short	rcv_wnd;		/* receive window */
	tcp_seq	rcv_nxt;		/* receive next */
	tcp_seq	rcv_up;			/* receive urgent pointer */
	tcp_seq	irs;			/* initial receive sequence number */
/*
 * Additional variables for this implementation.
 */
/* receive variables */
	tcp_seq	rcv_adv;		/* advertised window */
/* retransmit variables */
	tcp_seq	snd_max;		/* highest sequence number sent;
					 * used to recognize retransmits
					 */
/* congestion control (for slow start, source quench, retransmit after loss) */
	u_short	snd_cwnd;		/* congestion-controlled window */
	u_short snd_ssthresh;		/* snd_cwnd size threshhold for
					 * for slow start exponential to
					 * linear switch
					 */
/*
 * transmit timing stuff.  See below for scale of srtt and rttvar.
 * "Variance" is actually smoothed difference.
 */
	short	t_idle;			/* inactivity time */
	short	t_rtt;			/* round trip time */
	tcp_seq	t_rtseq;		/* sequence number being timed */
	short	t_srtt;			/* smoothed round-trip time */
	short	t_rttvar;		/* variance in round-trip time */
	u_short	t_rttmin;		/* minimum rtt allowed */
	u_short	max_sndwnd;		/* largest window peer has offered */

/* out-of-band data */
	char	t_oobflags;		/* have some */
	char	t_iobc;			/* input character */
#define	TCPOOB_HAVEDATA	0x01
#define	TCPOOB_HADDATA	0x02
	short	t_softerror;		/* possible error not yet reported */
};

#define	intotcpcb(ip)	((struct tcpcb *)(ip)->inp_ppcb)
#define	sototcpcb(so)	(intotcpcb(sotoinpcb(so)))

/*
 * The smoothed round-trip time and estimated variance
 * are stored as fixed point numbers scaled by the values below.
 * For convenience, these scales are also used in smoothing the average
 * (smoothed = (1/scale)sample + ((scale-1)/scale)smoothed).
 * With these scales, srtt has 3 bits to the right of the binary point,
 * and thus an "ALPHA" of 0.875.  rttvar has 2 bits to the right of the
 * binary point, and is smoothed with an ALPHA of 0.75.
 */
#define	TCP_RTT_SCALE		8	/* multiplier for srtt; 3 bits frac. */
#define	TCP_RTT_SHIFT		3	/* shift for srtt; 3 bits frac. */
#define	TCP_RTTVAR_SCALE	4	/* multiplier for rttvar; 2 bits */
#define	TCP_RTTVAR_SHIFT	2	/* multiplier for rttvar; 2 bits */

/*
 * The initial retransmission 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).
 * This macro assumes that the value of TCP_RTTVAR_SCALE
 * is the same as the multiplier for rttvar.
 */
#define	TCP_REXMTVAL(tp) \
	(((tp)->t_srtt >> TCP_RTT_SHIFT) + (tp)->t_rttvar)

/* XXX
 * We want to avoid doing m_pullup on incoming packets but that
 * means avoiding dtom on the tcp reassembly code.  That in turn means
 * keeping an mbuf pointer in the reassembly queue (since we might
 * have a cluster).  As a quick hack, the source & destination
 * port numbers (which are no longer needed once we've located the
 * tcpcb) are overlayed with an mbuf pointer.
 */
#define REASS_MBUF(ti) (*(struct mbuf **)&((ti)->ti_t))

/*
 * TCP statistics.
 * Many of these should be kept per connection,
 * but that's inconvenient at the moment.
 */
struct	tcpstat {
	u_long	tcps_connattempt;	/* connections initiated */
	u_long	tcps_accepts;		/* connections accepted */
	u_long	tcps_connects;		/* connections established */
	u_long	tcps_drops;		/* connections dropped */
	u_long	tcps_conndrops;		/* embryonic connections dropped */
	u_long	tcps_closed;		/* conn. closed (includes drops) */
	u_long	tcps_segstimed;		/* segs where we tried to get rtt */
	u_long	tcps_rttupdated;	/* times we succeeded */
	u_long	tcps_delack;		/* delayed acks sent */
	u_long	tcps_timeoutdrop;	/* conn. dropped in rxmt timeout */
	u_long	tcps_rexmttimeo;	/* retransmit timeouts */
	u_long	tcps_persisttimeo;	/* persist timeouts */
	u_long	tcps_keeptimeo;		/* keepalive timeouts */
	u_long	tcps_keepprobe;		/* keepalive probes sent */
	u_long	tcps_keepdrops;		/* connections dropped in keepalive */

	u_long	tcps_sndtotal;		/* total packets sent */
	u_long	tcps_sndpack;		/* data packets sent */
	u_long	tcps_sndbyte;		/* data bytes sent */
	u_long	tcps_sndrexmitpack;	/* data packets retransmitted */
	u_long	tcps_sndrexmitbyte;	/* data bytes retransmitted */
	u_long	tcps_sndacks;		/* ack-only packets sent */
	u_long	tcps_sndprobe;		/* window probes sent */
	u_long	tcps_sndurg;		/* packets sent with URG only */
	u_long	tcps_sndwinup;		/* window update-only packets sent */
	u_long	tcps_sndctrl;		/* control (SYN|FIN|RST) packets sent */

	u_long	tcps_rcvtotal;		/* total packets received */
	u_long	tcps_rcvpack;		/* packets received in sequence */
	u_long	tcps_rcvbyte;		/* bytes received in sequence */
	u_long	tcps_rcvbadsum;		/* packets received with ccksum errs */
	u_long	tcps_rcvbadoff;		/* packets received with bad offset */
	u_long	tcps_rcvshort;		/* packets received too short */
	u_long	tcps_rcvduppack;	/* duplicate-only packets received */
	u_long	tcps_rcvdupbyte;	/* duplicate-only bytes received */
	u_long	tcps_rcvpartduppack;	/* packets with some duplicate data */
	u_long	tcps_rcvpartdupbyte;	/* dup. bytes in part-dup. packets */
	u_long	tcps_rcvoopack;		/* out-of-order packets received */
	u_long	tcps_rcvoobyte;		/* out-of-order bytes received */
	u_long	tcps_rcvpackafterwin;	/* packets with data after window */
	u_long	tcps_rcvbyteafterwin;	/* bytes rcvd after window */
	u_long	tcps_rcvafterclose;	/* packets rcvd after "close" */
	u_long	tcps_rcvwinprobe;	/* rcvd window probe packets */
	u_long	tcps_rcvdupack;		/* rcvd duplicate acks */
	u_long	tcps_rcvacktoomuch;	/* rcvd acks for unsent data */
	u_long	tcps_rcvackpack;	/* rcvd ack packets */
	u_long	tcps_rcvackbyte;	/* bytes acked by rcvd acks */
	u_long	tcps_rcvwinupd;		/* rcvd window update packets */
};

#ifdef KERNEL
struct	inpcb tcb;		/* head of queue of active tcpcb's */
struct	tcpstat tcpstat;	/* tcp statistics */
struct	tcpiphdr *tcp_template();
struct	tcpcb *tcp_close(), *tcp_drop();
struct	tcpcb *tcp_timers(), *tcp_disconnect(), *tcp_usrclosed();
#endif
Commit	Line	Data
15637ed4 RG	1	/*
	2	* Copyright (c) 1982, 1986 Regents of the University of California.
	3	* 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	*
78ed81a3	33	* from: @(#)tcp_var.h 7.10 (Berkeley) 6/28/90
78ed81a3	34	* $Id$
15637ed4 RG	35	*/
	36
	37	/*
	38	* Kernel variables for tcp.
	39	*/
	40
	41	/*
	42	* Tcp control block, one per tcp; fields:
	43	*/
	44	struct tcpcb {
	45	struct tcpiphdr seg_next; / sequencing queue */
	46	struct tcpiphdr *seg_prev;
	47	short t_state; /* state of this connection */
	48	short t_timer[TCPT_NTIMERS]; /* tcp timers */
	49	short t_rxtshift; /* log(2) of rexmt exp. backoff */
	50	short t_rxtcur; /* current retransmit value */
	51	short t_dupacks; /* consecutive dup acks recd */
	52	u_short t_maxseg; /* maximum segment size */
	53	char t_force; /* 1 if forcing out a byte */
	54	u_char t_flags;
	55	#define TF_ACKNOW 0x01 /* ack peer immediately */
	56	#define TF_DELACK 0x02 /* ack, but try to delay it */
	57	#define TF_NODELAY 0x04 /* don't delay packets to coalesce */
	58	#define TF_NOOPT 0x08 /* don't use tcp options */
	59	#define TF_SENTFIN 0x10 /* have sent FIN */
	60	struct tcpiphdr t_template; / skeletal packet for transmit */
	61	struct inpcb t_inpcb; / back pointer to internet pcb */
	62	/*
	63	* The following fields are used as in the protocol specification.
	64	* See RFC783, Dec. 1981, page 21.
	65	*/
	66	/* send sequence variables */
	67	tcp_seq snd_una; /* send unacknowledged */
	68	tcp_seq snd_nxt; /* send next */
	69	tcp_seq snd_up; /* send urgent pointer */
	70	tcp_seq snd_wl1; /* window update seg seq number */
	71	tcp_seq snd_wl2; /* window update seg ack number */
	72	tcp_seq iss; /* initial send sequence number */
	73	u_short snd_wnd; /* send window */
	74	/* receive sequence variables */
	75	u_short rcv_wnd; /* receive window */
	76	tcp_seq rcv_nxt; /* receive next */
	77	tcp_seq rcv_up; /* receive urgent pointer */
	78	tcp_seq irs; /* initial receive sequence number */
	79	/*
	80	* Additional variables for this implementation.
	81	*/
	82	/* receive variables */
	83	tcp_seq rcv_adv; /* advertised window */
	84	/* retransmit variables */
	85	tcp_seq snd_max; /* highest sequence number sent;
	86	* used to recognize retransmits
	87	*/
	88	/* congestion control (for slow start, source quench, retransmit after loss) */
	89	u_short snd_cwnd; /* congestion-controlled window */
	90	u_short snd_ssthresh; /* snd_cwnd size threshhold for
	91	* for slow start exponential to
	92	* linear switch
	93	*/
	94	/*
	95	* transmit timing stuff. See below for scale of srtt and rttvar.
	96	* "Variance" is actually smoothed difference.
	97	*/
	98	short t_idle; /* inactivity time */
99	short t_rtt; /* round trip time */
100	tcp_seq t_rtseq; /* sequence number being timed */
101	short t_srtt; /* smoothed round-trip time */
102	short t_rttvar; /* variance in round-trip time */
103	u_short t_rttmin; /* minimum rtt allowed */
104	u_short max_sndwnd; /* largest window peer has offered */
105
106	/* out-of-band data */
107	char t_oobflags; /* have some */
108	char t_iobc; /* input character */
109	#define TCPOOB_HAVEDATA 0x01
110	#define TCPOOB_HADDATA 0x02
111	short t_softerror; /* possible error not yet reported */
112	};
113
114	#define intotcpcb(ip) ((struct tcpcb *)(ip)->inp_ppcb)
115	#define sototcpcb(so) (intotcpcb(sotoinpcb(so)))
116
117	/*
118	* The smoothed round-trip time and estimated variance
119	* are stored as fixed point numbers scaled by the values below.
120	* For convenience, these scales are also used in smoothing the average
121	* (smoothed = (1/scale)sample + ((scale-1)/scale)smoothed).
122	* With these scales, srtt has 3 bits to the right of the binary point,
123	* and thus an "ALPHA" of 0.875. rttvar has 2 bits to the right of the
124	* binary point, and is smoothed with an ALPHA of 0.75.
125	*/
126	#define TCP_RTT_SCALE 8 /* multiplier for srtt; 3 bits frac. */
127	#define TCP_RTT_SHIFT 3 /* shift for srtt; 3 bits frac. */
128	#define TCP_RTTVAR_SCALE 4 /* multiplier for rttvar; 2 bits */
129	#define TCP_RTTVAR_SHIFT 2 /* multiplier for rttvar; 2 bits */
130
131	/*
132	* The initial retransmission should happen at rtt + 4 * rttvar.
133	* Because of the way we do the smoothing, srtt and rttvar
134	* will each average +1/2 tick of bias. When we compute
135	* the retransmit timer, we want 1/2 tick of rounding and
136	* 1 extra tick because of +-1/2 tick uncertainty in the
137	* firing of the timer. The bias will give us exactly the
138	* 1.5 tick we need. But, because the bias is
139	* statistical, we have to test that we don't drop below
140	* the minimum feasible timer (which is 2 ticks).
141	* This macro assumes that the value of TCP_RTTVAR_SCALE
142	* is the same as the multiplier for rttvar.
143	*/
144	#define TCP_REXMTVAL(tp) \
145	(((tp)->t_srtt >> TCP_RTT_SHIFT) + (tp)->t_rttvar)
146
147	/* XXX
148	* We want to avoid doing m_pullup on incoming packets but that
149	* means avoiding dtom on the tcp reassembly code. That in turn means
150	* keeping an mbuf pointer in the reassembly queue (since we might
151	* have a cluster). As a quick hack, the source & destination
152	* port numbers (which are no longer needed once we've located the
153	* tcpcb) are overlayed with an mbuf pointer.
154	*/
155	#define REASS_MBUF(ti) ((struct mbuf *)&((ti)->ti_t))
156
157	/*
158	* TCP statistics.
159	* Many of these should be kept per connection,
160	* but that's inconvenient at the moment.
161	*/
162	struct tcpstat {
163	u_long tcps_connattempt; /* connections initiated */
164	u_long tcps_accepts; /* connections accepted */
165	u_long tcps_connects; /* connections established */
166	u_long tcps_drops; /* connections dropped */
167	u_long tcps_conndrops; /* embryonic connections dropped */
168	u_long tcps_closed; /* conn. closed (includes drops) */
169	u_long tcps_segstimed; /* segs where we tried to get rtt */
170	u_long tcps_rttupdated; /* times we succeeded */
171	u_long tcps_delack; /* delayed acks sent */
172	u_long tcps_timeoutdrop; /* conn. dropped in rxmt timeout */
173	u_long tcps_rexmttimeo; /* retransmit timeouts */
174	u_long tcps_persisttimeo; /* persist timeouts */
175	u_long tcps_keeptimeo; /* keepalive timeouts */
176	u_long tcps_keepprobe; /* keepalive probes sent */
177	u_long tcps_keepdrops; /* connections dropped in keepalive */
178
179	u_long tcps_sndtotal; /* total packets sent */
180	u_long tcps_sndpack; /* data packets sent */
181	u_long tcps_sndbyte; /* data bytes sent */
182	u_long tcps_sndrexmitpack; /* data packets retransmitted */
183	u_long tcps_sndrexmitbyte; /* data bytes retransmitted */
184	u_long tcps_sndacks; /* ack-only packets sent */
185	u_long tcps_sndprobe; /* window probes sent */
186	u_long tcps_sndurg; /* packets sent with URG only */
187	u_long tcps_sndwinup; /* window update-only packets sent */
188	u_long tcps_sndctrl; /* control (SYN\|FIN\|RST) packets sent */
189
190	u_long tcps_rcvtotal; /* total packets received */
191	u_long tcps_rcvpack; /* packets received in sequence */
192	u_long tcps_rcvbyte; /* bytes received in sequence */
193	u_long tcps_rcvbadsum; /* packets received with ccksum errs */
194	u_long tcps_rcvbadoff; /* packets received with bad offset */
195	u_long tcps_rcvshort; /* packets received too short */
196	u_long tcps_rcvduppack; /* duplicate-only packets received */
197	u_long tcps_rcvdupbyte; /* duplicate-only bytes received */
198	u_long tcps_rcvpartduppack; /* packets with some duplicate data */
199	u_long tcps_rcvpartdupbyte; /* dup. bytes in part-dup. packets */
200	u_long tcps_rcvoopack; /* out-of-order packets received */
201	u_long tcps_rcvoobyte; /* out-of-order bytes received */
202	u_long tcps_rcvpackafterwin; /* packets with data after window */
203	u_long tcps_rcvbyteafterwin; /* bytes rcvd after window */
204	u_long tcps_rcvafterclose; /* packets rcvd after "close" */
205	u_long tcps_rcvwinprobe; /* rcvd window probe packets */
206	u_long tcps_rcvdupack; /* rcvd duplicate acks */
207	u_long tcps_rcvacktoomuch; /* rcvd acks for unsent data */
208	u_long tcps_rcvackpack; /* rcvd ack packets */
209	u_long tcps_rcvackbyte; /* bytes acked by rcvd acks */
210	u_long tcps_rcvwinupd; /* rcvd window update packets */
211	};
212
213	#ifdef KERNEL
214	struct inpcb tcb; /* head of queue of active tcpcb's */
215	struct tcpstat tcpstat; /* tcp statistics */
216	struct tcpiphdr *tcp_template();
217	struct tcpcb tcp_close(), tcp_drop();
218	struct tcpcb tcp_timers(), tcp_disconnect(), *tcp_usrclosed();
219	#endif