1 /* $NetBSD: tcp_var.h,v 1.9 1994/06/29 06:38:54 cgd Exp $ */ 2 3 /* 4 * Copyright (c) 1982, 1986, 1993, 1994 5 * The Regents of the University of California. All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 3. All advertising materials mentioning features or use of this software 16 * must display the following acknowledgement: 17 * This product includes software developed by the University of 18 * California, Berkeley and its contributors. 19 * 4. Neither the name of the University nor the names of its contributors 20 * may be used to endorse or promote products derived from this software 21 * without specific prior written permission. 22 * 23 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 24 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 25 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 26 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 27 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 32 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 33 * SUCH DAMAGE. 34 * 35 * @(#)tcp_var.h 8.3 (Berkeley) 4/10/94 36 */ 37 38 /* 39 * Kernel variables for tcp. 40 */ 41 42 /* 43 * Tcp control block, one per tcp; fields: 44 */ 45 struct tcpcb { 46 struct tcpiphdr *seg_next; /* sequencing queue */ 47 struct tcpiphdr *seg_prev; 48 short t_state; /* state of this connection */ 49 short t_timer[TCPT_NTIMERS]; /* tcp timers */ 50 short t_rxtshift; /* log(2) of rexmt exp. backoff */ 51 short t_rxtcur; /* current retransmit value */ 52 short t_dupacks; /* consecutive dup acks recd */ 53 u_short t_maxseg; /* maximum segment size */ 54 char t_force; /* 1 if forcing out a byte */ 55 u_short t_flags; 56 #define TF_ACKNOW 0x0001 /* ack peer immediately */ 57 #define TF_DELACK 0x0002 /* ack, but try to delay it */ 58 #define TF_NODELAY 0x0004 /* don't delay packets to coalesce */ 59 #define TF_NOOPT 0x0008 /* don't use tcp options */ 60 #define TF_SENTFIN 0x0010 /* have sent FIN */ 61 #define TF_REQ_SCALE 0x0020 /* have/will request window scaling */ 62 #define TF_RCVD_SCALE 0x0040 /* other side has requested scaling */ 63 #define TF_REQ_TSTMP 0x0080 /* have/will request timestamps */ 64 #define TF_RCVD_TSTMP 0x0100 /* a timestamp was received in SYN */ 65 #define TF_SACK_PERMIT 0x0200 /* other side said I could SACK */ 66 67 struct tcpiphdr *t_template; /* skeletal packet for transmit */ 68 struct inpcb *t_inpcb; /* back pointer to internet pcb */ 69 /* 70 * The following fields are used as in the protocol specification. 71 * See RFC783, Dec. 1981, page 21. 72 */ 73 /* send sequence variables */ 74 tcp_seq snd_una; /* send unacknowledged */ 75 tcp_seq snd_nxt; /* send next */ 76 tcp_seq snd_up; /* send urgent pointer */ 77 tcp_seq snd_wl1; /* window update seg seq number */ 78 tcp_seq snd_wl2; /* window update seg ack number */ 79 tcp_seq iss; /* initial send sequence number */ 80 u_long snd_wnd; /* send window */ 81 /* receive sequence variables */ 82 u_long rcv_wnd; /* receive window */ 83 tcp_seq rcv_nxt; /* receive next */ 84 tcp_seq rcv_up; /* receive urgent pointer */ 85 tcp_seq irs; /* initial receive sequence number */ 86 /* 87 * Additional variables for this implementation. 88 */ 89 /* receive variables */ 90 tcp_seq rcv_adv; /* advertised window */ 91 /* retransmit variables */ 92 tcp_seq snd_max; /* highest sequence number sent; 93 * used to recognize retransmits 94 */ 95 /* congestion control (for slow start, source quench, retransmit after loss) */ 96 u_long snd_cwnd; /* congestion-controlled window */ 97 u_long snd_ssthresh; /* snd_cwnd size threshhold for 98 * for slow start exponential to 99 * linear switch 100 */ 101 /* 102 * transmit timing stuff. See below for scale of srtt and rttvar. 103 * "Variance" is actually smoothed difference. 104 */ 105 short t_idle; /* inactivity time */ 106 short t_rtt; /* round trip time */ 107 tcp_seq t_rtseq; /* sequence number being timed */ 108 short t_srtt; /* smoothed round-trip time */ 109 short t_rttvar; /* variance in round-trip time */ 110 u_short t_rttmin; /* minimum rtt allowed */ 111 u_long max_sndwnd; /* largest window peer has offered */ 112 113 /* out-of-band data */ 114 char t_oobflags; /* have some */ 115 char t_iobc; /* input character */ 116 #define TCPOOB_HAVEDATA 0x01 117 #define TCPOOB_HADDATA 0x02 118 short t_softerror; /* possible error not yet reported */ 119 120 /* RFC 1323 variables */ 121 u_char snd_scale; /* window scaling for send window */ 122 u_char rcv_scale; /* window scaling for recv window */ 123 u_char request_r_scale; /* pending window scaling */ 124 u_char requested_s_scale; 125 u_long ts_recent; /* timestamp echo data */ 126 u_long ts_recent_age; /* when last updated */ 127 tcp_seq last_ack_sent; 128 129 /* TUBA stuff */ 130 caddr_t t_tuba_pcb; /* next level down pcb for TCP over z */ 131 }; 132 133 #define intotcpcb(ip) ((struct tcpcb *)(ip)->inp_ppcb) 134 #define sototcpcb(so) (intotcpcb(sotoinpcb(so))) 135 136 /* 137 * The smoothed round-trip time and estimated variance 138 * are stored as fixed point numbers scaled by the values below. 139 * For convenience, these scales are also used in smoothing the average 140 * (smoothed = (1/scale)sample + ((scale-1)/scale)smoothed). 141 * With these scales, srtt has 3 bits to the right of the binary point, 142 * and thus an "ALPHA" of 0.875. rttvar has 2 bits to the right of the 143 * binary point, and is smoothed with an ALPHA of 0.75. 144 */ 145 #define TCP_RTT_SCALE 8 /* multiplier for srtt; 3 bits frac. */ 146 #define TCP_RTT_SHIFT 3 /* shift for srtt; 3 bits frac. */ 147 #define TCP_RTTVAR_SCALE 4 /* multiplier for rttvar; 2 bits */ 148 #define TCP_RTTVAR_SHIFT 2 /* multiplier for rttvar; 2 bits */ 149 150 /* 151 * The initial retransmission should happen at rtt + 4 * rttvar. 152 * Because of the way we do the smoothing, srtt and rttvar 153 * will each average +1/2 tick of bias. When we compute 154 * the retransmit timer, we want 1/2 tick of rounding and 155 * 1 extra tick because of +-1/2 tick uncertainty in the 156 * firing of the timer. The bias will give us exactly the 157 * 1.5 tick we need. But, because the bias is 158 * statistical, we have to test that we don't drop below 159 * the minimum feasible timer (which is 2 ticks). 160 * This macro assumes that the value of TCP_RTTVAR_SCALE 161 * is the same as the multiplier for rttvar. 162 */ 163 #define TCP_REXMTVAL(tp) \ 164 (((tp)->t_srtt >> TCP_RTT_SHIFT) + (tp)->t_rttvar) 165 166 /* XXX 167 * We want to avoid doing m_pullup on incoming packets but that 168 * means avoiding dtom on the tcp reassembly code. That in turn means 169 * keeping an mbuf pointer in the reassembly queue (since we might 170 * have a cluster). As a quick hack, the source & destination 171 * port numbers (which are no longer needed once we've located the 172 * tcpcb) are overlayed with an mbuf pointer. 173 */ 174 #define REASS_MBUF(ti) (*(struct mbuf **)&((ti)->ti_t)) 175 176 /* 177 * TCP statistics. 178 * Many of these should be kept per connection, 179 * but that's inconvenient at the moment. 180 */ 181 struct tcpstat { 182 u_long tcps_connattempt; /* connections initiated */ 183 u_long tcps_accepts; /* connections accepted */ 184 u_long tcps_connects; /* connections established */ 185 u_long tcps_drops; /* connections dropped */ 186 u_long tcps_conndrops; /* embryonic connections dropped */ 187 u_long tcps_closed; /* conn. closed (includes drops) */ 188 u_long tcps_segstimed; /* segs where we tried to get rtt */ 189 u_long tcps_rttupdated; /* times we succeeded */ 190 u_long tcps_delack; /* delayed acks sent */ 191 u_long tcps_timeoutdrop; /* conn. dropped in rxmt timeout */ 192 u_long tcps_rexmttimeo; /* retransmit timeouts */ 193 u_long tcps_persisttimeo; /* persist timeouts */ 194 u_long tcps_keeptimeo; /* keepalive timeouts */ 195 u_long tcps_keepprobe; /* keepalive probes sent */ 196 u_long tcps_keepdrops; /* connections dropped in keepalive */ 197 198 u_long tcps_sndtotal; /* total packets sent */ 199 u_long tcps_sndpack; /* data packets sent */ 200 u_long tcps_sndbyte; /* data bytes sent */ 201 u_long tcps_sndrexmitpack; /* data packets retransmitted */ 202 u_long tcps_sndrexmitbyte; /* data bytes retransmitted */ 203 u_long tcps_sndacks; /* ack-only packets sent */ 204 u_long tcps_sndprobe; /* window probes sent */ 205 u_long tcps_sndurg; /* packets sent with URG only */ 206 u_long tcps_sndwinup; /* window update-only packets sent */ 207 u_long tcps_sndctrl; /* control (SYN|FIN|RST) packets sent */ 208 209 u_long tcps_rcvtotal; /* total packets received */ 210 u_long tcps_rcvpack; /* packets received in sequence */ 211 u_long tcps_rcvbyte; /* bytes received in sequence */ 212 u_long tcps_rcvbadsum; /* packets received with ccksum errs */ 213 u_long tcps_rcvbadoff; /* packets received with bad offset */ 214 u_long tcps_rcvshort; /* packets received too short */ 215 u_long tcps_rcvduppack; /* duplicate-only packets received */ 216 u_long tcps_rcvdupbyte; /* duplicate-only bytes received */ 217 u_long tcps_rcvpartduppack; /* packets with some duplicate data */ 218 u_long tcps_rcvpartdupbyte; /* dup. bytes in part-dup. packets */ 219 u_long tcps_rcvoopack; /* out-of-order packets received */ 220 u_long tcps_rcvoobyte; /* out-of-order bytes received */ 221 u_long tcps_rcvpackafterwin; /* packets with data after window */ 222 u_long tcps_rcvbyteafterwin; /* bytes rcvd after window */ 223 u_long tcps_rcvafterclose; /* packets rcvd after "close" */ 224 u_long tcps_rcvwinprobe; /* rcvd window probe packets */ 225 u_long tcps_rcvdupack; /* rcvd duplicate acks */ 226 u_long tcps_rcvacktoomuch; /* rcvd acks for unsent data */ 227 u_long tcps_rcvackpack; /* rcvd ack packets */ 228 u_long tcps_rcvackbyte; /* bytes acked by rcvd acks */ 229 u_long tcps_rcvwinupd; /* rcvd window update packets */ 230 u_long tcps_pawsdrop; /* segments dropped due to PAWS */ 231 u_long tcps_predack; /* times hdr predict ok for acks */ 232 u_long tcps_preddat; /* times hdr predict ok for data pkts */ 233 u_long tcps_pcbcachemiss; 234 }; 235 236 #ifdef KERNEL 237 struct inpcb tcb; /* head of queue of active tcpcb's */ 238 struct tcpstat tcpstat; /* tcp statistics */ 239 u_long tcp_now; /* for RFC 1323 timestamps */ 240 241 int tcp_attach __P((struct socket *)); 242 void tcp_canceltimers __P((struct tcpcb *)); 243 struct tcpcb * 244 tcp_close __P((struct tcpcb *)); 245 void tcp_ctlinput __P((int, struct sockaddr *, struct ip *)); 246 int tcp_ctloutput __P((int, struct socket *, int, int, struct mbuf **)); 247 struct tcpcb * 248 tcp_disconnect __P((struct tcpcb *)); 249 struct tcpcb * 250 tcp_drop __P((struct tcpcb *, int)); 251 void tcp_dooptions __P((struct tcpcb *, 252 u_char *, int, struct tcpiphdr *, int *, u_long *, u_long *)); 253 void tcp_drain __P((void)); 254 void tcp_fasttimo __P((void)); 255 void tcp_init __P((void)); 256 void tcp_input __P((struct mbuf *, int)); 257 int tcp_mss __P((struct tcpcb *, u_int)); 258 struct tcpcb * 259 tcp_newtcpcb __P((struct inpcb *)); 260 void tcp_notify __P((struct inpcb *, int)); 261 int tcp_output __P((struct tcpcb *)); 262 void tcp_pulloutofband __P((struct socket *, 263 struct tcpiphdr *, struct mbuf *)); 264 void tcp_quench __P((struct inpcb *, int)); 265 int tcp_reass __P((struct tcpcb *, struct tcpiphdr *, struct mbuf *)); 266 void tcp_respond __P((struct tcpcb *, 267 struct tcpiphdr *, struct mbuf *, u_long, u_long, int)); 268 void tcp_setpersist __P((struct tcpcb *)); 269 void tcp_slowtimo __P((void)); 270 struct tcpiphdr * 271 tcp_template __P((struct tcpcb *)); 272 struct tcpcb * 273 tcp_timers __P((struct tcpcb *, int)); 274 void tcp_trace __P((int, int, struct tcpcb *, struct tcpiphdr *, int)); 275 struct tcpcb * 276 tcp_usrclosed __P((struct tcpcb *)); 277 int tcp_usrreq __P((struct socket *, 278 int, struct mbuf *, struct mbuf *, struct mbuf *)); 279 void tcp_xmit_timer __P((struct tcpcb *, int)); 280 #endif 281