The Design and Implementation of the FreeBSD Operating System, Second Edition
Now available: The Design and Implementation of the FreeBSD Operating System (Second Edition)


[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]

FreeBSD/Linux Kernel Cross Reference
sys/netinet/tcp_sack.c

Version: -  FREEBSD  -  FREEBSD-13-STABLE  -  FREEBSD-13-0  -  FREEBSD-12-STABLE  -  FREEBSD-12-0  -  FREEBSD-11-STABLE  -  FREEBSD-11-0  -  FREEBSD-10-STABLE  -  FREEBSD-10-0  -  FREEBSD-9-STABLE  -  FREEBSD-9-0  -  FREEBSD-8-STABLE  -  FREEBSD-8-0  -  FREEBSD-7-STABLE  -  FREEBSD-7-0  -  FREEBSD-6-STABLE  -  FREEBSD-6-0  -  FREEBSD-5-STABLE  -  FREEBSD-5-0  -  FREEBSD-4-STABLE  -  FREEBSD-3-STABLE  -  FREEBSD22  -  l41  -  OPENBSD  -  linux-2.6  -  MK84  -  PLAN9  -  xnu-8792 
SearchContext: -  none  -  3  -  10 

    1 /*-
    2  * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1994, 1995
    3  *      The Regents of the University of California.
    4  * All rights reserved.
    5  *
    6  * Redistribution and use in source and binary forms, with or without
    7  * modification, are permitted provided that the following conditions
    8  * are met:
    9  * 1. Redistributions of source code must retain the above copyright
   10  *    notice, this list of conditions and the following disclaimer.
   11  * 2. Redistributions in binary form must reproduce the above copyright
   12  *    notice, this list of conditions and the following disclaimer in the
   13  *    documentation and/or other materials provided with the distribution.
   14  * 4. Neither the name of the University nor the names of its contributors
   15  *    may be used to endorse or promote products derived from this software
   16  *    without specific prior written permission.
   17  *
   18  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
   19  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
   20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
   21  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
   22  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
   23  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
   24  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
   25  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
   26  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
   27  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
   28  * SUCH DAMAGE.
   29  *
   30  *      @(#)tcp_sack.c  8.12 (Berkeley) 5/24/95
   31  */
   32 
   33 /*-
   34  *      @@(#)COPYRIGHT  1.1 (NRL) 17 January 1995
   35  *
   36  * NRL grants permission for redistribution and use in source and binary
   37  * forms, with or without modification, of the software and documentation
   38  * created at NRL provided that the following conditions are met:
   39  *
   40  * 1. Redistributions of source code must retain the above copyright
   41  *    notice, this list of conditions and the following disclaimer.
   42  * 2. Redistributions in binary form must reproduce the above copyright
   43  *    notice, this list of conditions and the following disclaimer in the
   44  *    documentation and/or other materials provided with the distribution.
   45  * 3. All advertising materials mentioning features or use of this software
   46  *    must display the following acknowledgements:
   47  *      This product includes software developed by the University of
   48  *      California, Berkeley and its contributors.
   49  *      This product includes software developed at the Information
   50  *      Technology Division, US Naval Research Laboratory.
   51  * 4. Neither the name of the NRL nor the names of its contributors
   52  *    may be used to endorse or promote products derived from this software
   53  *    without specific prior written permission.
   54  *
   55  * THE SOFTWARE PROVIDED BY NRL IS PROVIDED BY NRL AND CONTRIBUTORS ``AS
   56  * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
   57  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
   58  * PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL NRL OR
   59  * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
   60  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
   61  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
   62  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
   63  * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
   64  * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
   65  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
   66  *
   67  * The views and conclusions contained in the software and documentation
   68  * are those of the authors and should not be interpreted as representing
   69  * official policies, either expressed or implied, of the US Naval
   70  * Research Laboratory (NRL).
   71  */
   72 
   73 #include <sys/cdefs.h>
   74 __FBSDID("$FreeBSD: releng/9.2/sys/netinet/tcp_sack.c 220105 2011-03-28 19:03:56Z weongyo $");
   75 
   76 #include "opt_inet.h"
   77 #include "opt_inet6.h"
   78 #include "opt_tcpdebug.h"
   79 
   80 #include <sys/param.h>
   81 #include <sys/systm.h>
   82 #include <sys/kernel.h>
   83 #include <sys/sysctl.h>
   84 #include <sys/malloc.h>
   85 #include <sys/mbuf.h>
   86 #include <sys/proc.h>           /* for proc0 declaration */
   87 #include <sys/protosw.h>
   88 #include <sys/socket.h>
   89 #include <sys/socketvar.h>
   90 #include <sys/syslog.h>
   91 #include <sys/systm.h>
   92 
   93 #include <machine/cpu.h>        /* before tcp_seq.h, for tcp_random18() */
   94 
   95 #include <vm/uma.h>
   96 
   97 #include <net/if.h>
   98 #include <net/route.h>
   99 #include <net/vnet.h>
  100 
  101 #include <netinet/in.h>
  102 #include <netinet/in_systm.h>
  103 #include <netinet/ip.h>
  104 #include <netinet/in_var.h>
  105 #include <netinet/in_pcb.h>
  106 #include <netinet/ip_var.h>
  107 #include <netinet/ip6.h>
  108 #include <netinet/icmp6.h>
  109 #include <netinet6/nd6.h>
  110 #include <netinet6/ip6_var.h>
  111 #include <netinet6/in6_pcb.h>
  112 #include <netinet/tcp.h>
  113 #include <netinet/tcp_fsm.h>
  114 #include <netinet/tcp_seq.h>
  115 #include <netinet/tcp_timer.h>
  116 #include <netinet/tcp_var.h>
  117 #include <netinet6/tcp6_var.h>
  118 #include <netinet/tcpip.h>
  119 #ifdef TCPDEBUG
  120 #include <netinet/tcp_debug.h>
  121 #endif /* TCPDEBUG */
  122 
  123 #include <machine/in_cksum.h>
  124 
  125 VNET_DECLARE(struct uma_zone *, sack_hole_zone);
  126 #define V_sack_hole_zone                VNET(sack_hole_zone)
  127 
  128 SYSCTL_NODE(_net_inet_tcp, OID_AUTO, sack, CTLFLAG_RW, 0, "TCP SACK");
  129 VNET_DEFINE(int, tcp_do_sack) = 1;
  130 #define V_tcp_do_sack                   VNET(tcp_do_sack)
  131 SYSCTL_VNET_INT(_net_inet_tcp_sack, OID_AUTO, enable, CTLFLAG_RW,
  132     &VNET_NAME(tcp_do_sack), 0, "Enable/Disable TCP SACK support");
  133 
  134 VNET_DEFINE(int, tcp_sack_maxholes) = 128;
  135 #define V_tcp_sack_maxholes             VNET(tcp_sack_maxholes)
  136 SYSCTL_VNET_INT(_net_inet_tcp_sack, OID_AUTO, maxholes, CTLFLAG_RW,
  137     &VNET_NAME(tcp_sack_maxholes), 0,
  138     "Maximum number of TCP SACK holes allowed per connection");
  139 
  140 VNET_DEFINE(int, tcp_sack_globalmaxholes) = 65536;
  141 #define V_tcp_sack_globalmaxholes       VNET(tcp_sack_globalmaxholes)
  142 SYSCTL_VNET_INT(_net_inet_tcp_sack, OID_AUTO, globalmaxholes, CTLFLAG_RW,
  143     &VNET_NAME(tcp_sack_globalmaxholes), 0, 
  144     "Global maximum number of TCP SACK holes");
  145 
  146 VNET_DEFINE(int, tcp_sack_globalholes) = 0;
  147 #define V_tcp_sack_globalholes          VNET(tcp_sack_globalholes)
  148 SYSCTL_VNET_INT(_net_inet_tcp_sack, OID_AUTO, globalholes, CTLFLAG_RD,
  149     &VNET_NAME(tcp_sack_globalholes), 0,
  150     "Global number of TCP SACK holes currently allocated");
  151 
  152 /*
  153  * This function is called upon receipt of new valid data (while not in
  154  * header prediction mode), and it updates the ordered list of sacks.
  155  */
  156 void
  157 tcp_update_sack_list(struct tcpcb *tp, tcp_seq rcv_start, tcp_seq rcv_end)
  158 {
  159         /*
  160          * First reported block MUST be the most recent one.  Subsequent
  161          * blocks SHOULD be in the order in which they arrived at the
  162          * receiver.  These two conditions make the implementation fully
  163          * compliant with RFC 2018.
  164          */
  165         struct sackblk head_blk, saved_blks[MAX_SACK_BLKS];
  166         int num_head, num_saved, i;
  167 
  168         INP_WLOCK_ASSERT(tp->t_inpcb);
  169 
  170         /* Check arguments. */
  171         KASSERT(SEQ_LT(rcv_start, rcv_end), ("rcv_start < rcv_end"));
  172 
  173         /* SACK block for the received segment. */
  174         head_blk.start = rcv_start;
  175         head_blk.end = rcv_end;
  176 
  177         /*
  178          * Merge updated SACK blocks into head_blk, and save unchanged SACK
  179          * blocks into saved_blks[].  num_saved will have the number of the
  180          * saved SACK blocks.
  181          */
  182         num_saved = 0;
  183         for (i = 0; i < tp->rcv_numsacks; i++) {
  184                 tcp_seq start = tp->sackblks[i].start;
  185                 tcp_seq end = tp->sackblks[i].end;
  186                 if (SEQ_GEQ(start, end) || SEQ_LEQ(start, tp->rcv_nxt)) {
  187                         /*
  188                          * Discard this SACK block.
  189                          */
  190                 } else if (SEQ_LEQ(head_blk.start, end) &&
  191                            SEQ_GEQ(head_blk.end, start)) {
  192                         /*
  193                          * Merge this SACK block into head_blk.  This SACK
  194                          * block itself will be discarded.
  195                          */
  196                         if (SEQ_GT(head_blk.start, start))
  197                                 head_blk.start = start;
  198                         if (SEQ_LT(head_blk.end, end))
  199                                 head_blk.end = end;
  200                 } else {
  201                         /*
  202                          * Save this SACK block.
  203                          */
  204                         saved_blks[num_saved].start = start;
  205                         saved_blks[num_saved].end = end;
  206                         num_saved++;
  207                 }
  208         }
  209 
  210         /*
  211          * Update SACK list in tp->sackblks[].
  212          */
  213         num_head = 0;
  214         if (SEQ_GT(head_blk.start, tp->rcv_nxt)) {
  215                 /*
  216                  * The received data segment is an out-of-order segment.  Put
  217                  * head_blk at the top of SACK list.
  218                  */
  219                 tp->sackblks[0] = head_blk;
  220                 num_head = 1;
  221                 /*
  222                  * If the number of saved SACK blocks exceeds its limit,
  223                  * discard the last SACK block.
  224                  */
  225                 if (num_saved >= MAX_SACK_BLKS)
  226                         num_saved--;
  227         }
  228         if (num_saved > 0) {
  229                 /*
  230                  * Copy the saved SACK blocks back.
  231                  */
  232                 bcopy(saved_blks, &tp->sackblks[num_head],
  233                       sizeof(struct sackblk) * num_saved);
  234         }
  235 
  236         /* Save the number of SACK blocks. */
  237         tp->rcv_numsacks = num_head + num_saved;
  238 }
  239 
  240 /*
  241  * Delete all receiver-side SACK information.
  242  */
  243 void
  244 tcp_clean_sackreport(struct tcpcb *tp)
  245 {
  246         int i;
  247 
  248         INP_WLOCK_ASSERT(tp->t_inpcb);
  249         tp->rcv_numsacks = 0;
  250         for (i = 0; i < MAX_SACK_BLKS; i++)
  251                 tp->sackblks[i].start = tp->sackblks[i].end=0;
  252 }
  253 
  254 /*
  255  * Allocate struct sackhole.
  256  */
  257 static struct sackhole *
  258 tcp_sackhole_alloc(struct tcpcb *tp, tcp_seq start, tcp_seq end)
  259 {
  260         struct sackhole *hole;
  261 
  262         if (tp->snd_numholes >= V_tcp_sack_maxholes ||
  263             V_tcp_sack_globalholes >= V_tcp_sack_globalmaxholes) {
  264                 TCPSTAT_INC(tcps_sack_sboverflow);
  265                 return NULL;
  266         }
  267 
  268         hole = (struct sackhole *)uma_zalloc(V_sack_hole_zone, M_NOWAIT);
  269         if (hole == NULL)
  270                 return NULL;
  271 
  272         hole->start = start;
  273         hole->end = end;
  274         hole->rxmit = start;
  275 
  276         tp->snd_numholes++;
  277         atomic_add_int(&V_tcp_sack_globalholes, 1);
  278 
  279         return hole;
  280 }
  281 
  282 /*
  283  * Free struct sackhole.
  284  */
  285 static void
  286 tcp_sackhole_free(struct tcpcb *tp, struct sackhole *hole)
  287 {
  288 
  289         uma_zfree(V_sack_hole_zone, hole);
  290 
  291         tp->snd_numholes--;
  292         atomic_subtract_int(&V_tcp_sack_globalholes, 1);
  293 
  294         KASSERT(tp->snd_numholes >= 0, ("tp->snd_numholes >= 0"));
  295         KASSERT(V_tcp_sack_globalholes >= 0, ("tcp_sack_globalholes >= 0"));
  296 }
  297 
  298 /*
  299  * Insert new SACK hole into scoreboard.
  300  */
  301 static struct sackhole *
  302 tcp_sackhole_insert(struct tcpcb *tp, tcp_seq start, tcp_seq end,
  303     struct sackhole *after)
  304 {
  305         struct sackhole *hole;
  306 
  307         /* Allocate a new SACK hole. */
  308         hole = tcp_sackhole_alloc(tp, start, end);
  309         if (hole == NULL)
  310                 return NULL;
  311 
  312         /* Insert the new SACK hole into scoreboard. */
  313         if (after != NULL)
  314                 TAILQ_INSERT_AFTER(&tp->snd_holes, after, hole, scblink);
  315         else
  316                 TAILQ_INSERT_TAIL(&tp->snd_holes, hole, scblink);
  317 
  318         /* Update SACK hint. */
  319         if (tp->sackhint.nexthole == NULL)
  320                 tp->sackhint.nexthole = hole;
  321 
  322         return hole;
  323 }
  324 
  325 /*
  326  * Remove SACK hole from scoreboard.
  327  */
  328 static void
  329 tcp_sackhole_remove(struct tcpcb *tp, struct sackhole *hole)
  330 {
  331 
  332         /* Update SACK hint. */
  333         if (tp->sackhint.nexthole == hole)
  334                 tp->sackhint.nexthole = TAILQ_NEXT(hole, scblink);
  335 
  336         /* Remove this SACK hole. */
  337         TAILQ_REMOVE(&tp->snd_holes, hole, scblink);
  338 
  339         /* Free this SACK hole. */
  340         tcp_sackhole_free(tp, hole);
  341 }
  342 
  343 /*
  344  * Process cumulative ACK and the TCP SACK option to update the scoreboard.
  345  * tp->snd_holes is an ordered list of holes (oldest to newest, in terms of
  346  * the sequence space).
  347  */
  348 void
  349 tcp_sack_doack(struct tcpcb *tp, struct tcpopt *to, tcp_seq th_ack)
  350 {
  351         struct sackhole *cur, *temp;
  352         struct sackblk sack, sack_blocks[TCP_MAX_SACK + 1], *sblkp;
  353         int i, j, num_sack_blks;
  354 
  355         INP_WLOCK_ASSERT(tp->t_inpcb);
  356 
  357         num_sack_blks = 0;
  358         /*
  359          * If SND.UNA will be advanced by SEG.ACK, and if SACK holes exist,
  360          * treat [SND.UNA, SEG.ACK) as if it is a SACK block.
  361          */
  362         if (SEQ_LT(tp->snd_una, th_ack) && !TAILQ_EMPTY(&tp->snd_holes)) {
  363                 sack_blocks[num_sack_blks].start = tp->snd_una;
  364                 sack_blocks[num_sack_blks++].end = th_ack;
  365         }
  366         /*
  367          * Append received valid SACK blocks to sack_blocks[], but only if we
  368          * received new blocks from the other side.
  369          */
  370         if (to->to_flags & TOF_SACK) {
  371                 for (i = 0; i < to->to_nsacks; i++) {
  372                         bcopy((to->to_sacks + i * TCPOLEN_SACK),
  373                             &sack, sizeof(sack));
  374                         sack.start = ntohl(sack.start);
  375                         sack.end = ntohl(sack.end);
  376                         if (SEQ_GT(sack.end, sack.start) &&
  377                             SEQ_GT(sack.start, tp->snd_una) &&
  378                             SEQ_GT(sack.start, th_ack) &&
  379                             SEQ_LT(sack.start, tp->snd_max) &&
  380                             SEQ_GT(sack.end, tp->snd_una) &&
  381                             SEQ_LEQ(sack.end, tp->snd_max))
  382                                 sack_blocks[num_sack_blks++] = sack;
  383                 }
  384         }
  385         /*
  386          * Return if SND.UNA is not advanced and no valid SACK block is
  387          * received.
  388          */
  389         if (num_sack_blks == 0)
  390                 return;
  391 
  392         /*
  393          * Sort the SACK blocks so we can update the scoreboard with just one
  394          * pass. The overhead of sorting upto 4+1 elements is less than
  395          * making upto 4+1 passes over the scoreboard.
  396          */
  397         for (i = 0; i < num_sack_blks; i++) {
  398                 for (j = i + 1; j < num_sack_blks; j++) {
  399                         if (SEQ_GT(sack_blocks[i].end, sack_blocks[j].end)) {
  400                                 sack = sack_blocks[i];
  401                                 sack_blocks[i] = sack_blocks[j];
  402                                 sack_blocks[j] = sack;
  403                         }
  404                 }
  405         }
  406         if (TAILQ_EMPTY(&tp->snd_holes))
  407                 /*
  408                  * Empty scoreboard. Need to initialize snd_fack (it may be
  409                  * uninitialized or have a bogus value). Scoreboard holes
  410                  * (from the sack blocks received) are created later below
  411                  * (in the logic that adds holes to the tail of the
  412                  * scoreboard).
  413                  */
  414                 tp->snd_fack = SEQ_MAX(tp->snd_una, th_ack);
  415         /*
  416          * In the while-loop below, incoming SACK blocks (sack_blocks[]) and
  417          * SACK holes (snd_holes) are traversed from their tails with just
  418          * one pass in order to reduce the number of compares especially when
  419          * the bandwidth-delay product is large.
  420          *
  421          * Note: Typically, in the first RTT of SACK recovery, the highest
  422          * three or four SACK blocks with the same ack number are received.
  423          * In the second RTT, if retransmitted data segments are not lost,
  424          * the highest three or four SACK blocks with ack number advancing
  425          * are received.
  426          */
  427         sblkp = &sack_blocks[num_sack_blks - 1];        /* Last SACK block */
  428         tp->sackhint.last_sack_ack = sblkp->end;
  429         if (SEQ_LT(tp->snd_fack, sblkp->start)) {
  430                 /*
  431                  * The highest SACK block is beyond fack.  Append new SACK
  432                  * hole at the tail.  If the second or later highest SACK
  433                  * blocks are also beyond the current fack, they will be
  434                  * inserted by way of hole splitting in the while-loop below.
  435                  */
  436                 temp = tcp_sackhole_insert(tp, tp->snd_fack,sblkp->start,NULL);
  437                 if (temp != NULL) {
  438                         tp->snd_fack = sblkp->end;
  439                         /* Go to the previous sack block. */
  440                         sblkp--;
  441                 } else {
  442                         /* 
  443                          * We failed to add a new hole based on the current 
  444                          * sack block.  Skip over all the sack blocks that 
  445                          * fall completely to the right of snd_fack and
  446                          * proceed to trim the scoreboard based on the
  447                          * remaining sack blocks.  This also trims the
  448                          * scoreboard for th_ack (which is sack_blocks[0]).
  449                          */
  450                         while (sblkp >= sack_blocks && 
  451                                SEQ_LT(tp->snd_fack, sblkp->start))
  452                                 sblkp--;
  453                         if (sblkp >= sack_blocks && 
  454                             SEQ_LT(tp->snd_fack, sblkp->end))
  455                                 tp->snd_fack = sblkp->end;
  456                 }
  457         } else if (SEQ_LT(tp->snd_fack, sblkp->end))
  458                 /* fack is advanced. */
  459                 tp->snd_fack = sblkp->end;
  460         /* We must have at least one SACK hole in scoreboard. */
  461         KASSERT(!TAILQ_EMPTY(&tp->snd_holes),
  462             ("SACK scoreboard must not be empty"));
  463         cur = TAILQ_LAST(&tp->snd_holes, sackhole_head); /* Last SACK hole. */
  464         /*
  465          * Since the incoming sack blocks are sorted, we can process them
  466          * making one sweep of the scoreboard.
  467          */
  468         while (sblkp >= sack_blocks  && cur != NULL) {
  469                 if (SEQ_GEQ(sblkp->start, cur->end)) {
  470                         /*
  471                          * SACKs data beyond the current hole.  Go to the
  472                          * previous sack block.
  473                          */
  474                         sblkp--;
  475                         continue;
  476                 }
  477                 if (SEQ_LEQ(sblkp->end, cur->start)) {
  478                         /*
  479                          * SACKs data before the current hole.  Go to the
  480                          * previous hole.
  481                          */
  482                         cur = TAILQ_PREV(cur, sackhole_head, scblink);
  483                         continue;
  484                 }
  485                 tp->sackhint.sack_bytes_rexmit -= (cur->rxmit - cur->start);
  486                 KASSERT(tp->sackhint.sack_bytes_rexmit >= 0,
  487                     ("sackhint bytes rtx >= 0"));
  488                 if (SEQ_LEQ(sblkp->start, cur->start)) {
  489                         /* Data acks at least the beginning of hole. */
  490                         if (SEQ_GEQ(sblkp->end, cur->end)) {
  491                                 /* Acks entire hole, so delete hole. */
  492                                 temp = cur;
  493                                 cur = TAILQ_PREV(cur, sackhole_head, scblink);
  494                                 tcp_sackhole_remove(tp, temp);
  495                                 /*
  496                                  * The sack block may ack all or part of the
  497                                  * next hole too, so continue onto the next
  498                                  * hole.
  499                                  */
  500                                 continue;
  501                         } else {
  502                                 /* Move start of hole forward. */
  503                                 cur->start = sblkp->end;
  504                                 cur->rxmit = SEQ_MAX(cur->rxmit, cur->start);
  505                         }
  506                 } else {
  507                         /* Data acks at least the end of hole. */
  508                         if (SEQ_GEQ(sblkp->end, cur->end)) {
  509                                 /* Move end of hole backward. */
  510                                 cur->end = sblkp->start;
  511                                 cur->rxmit = SEQ_MIN(cur->rxmit, cur->end);
  512                         } else {
  513                                 /*
  514                                  * ACKs some data in middle of a hole; need
  515                                  * to split current hole
  516                                  */
  517                                 temp = tcp_sackhole_insert(tp, sblkp->end,
  518                                     cur->end, cur);
  519                                 if (temp != NULL) {
  520                                         if (SEQ_GT(cur->rxmit, temp->rxmit)) {
  521                                                 temp->rxmit = cur->rxmit;
  522                                                 tp->sackhint.sack_bytes_rexmit
  523                                                     += (temp->rxmit
  524                                                     - temp->start);
  525                                         }
  526                                         cur->end = sblkp->start;
  527                                         cur->rxmit = SEQ_MIN(cur->rxmit,
  528                                             cur->end);
  529                                 }
  530                         }
  531                 }
  532                 tp->sackhint.sack_bytes_rexmit += (cur->rxmit - cur->start);
  533                 /*
  534                  * Testing sblkp->start against cur->start tells us whether
  535                  * we're done with the sack block or the sack hole.
  536                  * Accordingly, we advance one or the other.
  537                  */
  538                 if (SEQ_LEQ(sblkp->start, cur->start))
  539                         cur = TAILQ_PREV(cur, sackhole_head, scblink);
  540                 else
  541                         sblkp--;
  542         }
  543 }
  544 
  545 /*
  546  * Free all SACK holes to clear the scoreboard.
  547  */
  548 void
  549 tcp_free_sackholes(struct tcpcb *tp)
  550 {
  551         struct sackhole *q;
  552 
  553         INP_WLOCK_ASSERT(tp->t_inpcb);
  554         while ((q = TAILQ_FIRST(&tp->snd_holes)) != NULL)
  555                 tcp_sackhole_remove(tp, q);
  556         tp->sackhint.sack_bytes_rexmit = 0;
  557 
  558         KASSERT(tp->snd_numholes == 0, ("tp->snd_numholes == 0"));
  559         KASSERT(tp->sackhint.nexthole == NULL,
  560                 ("tp->sackhint.nexthole == NULL"));
  561 }
  562 
  563 /*
  564  * Partial ack handling within a sack recovery episode.  Keeping this very
  565  * simple for now.  When a partial ack is received, force snd_cwnd to a value
  566  * that will allow the sender to transmit no more than 2 segments.  If
  567  * necessary, a better scheme can be adopted at a later point, but for now,
  568  * the goal is to prevent the sender from bursting a large amount of data in
  569  * the midst of sack recovery.
  570  */
  571 void
  572 tcp_sack_partialack(struct tcpcb *tp, struct tcphdr *th)
  573 {
  574         int num_segs = 1;
  575 
  576         INP_WLOCK_ASSERT(tp->t_inpcb);
  577         tcp_timer_activate(tp, TT_REXMT, 0);
  578         tp->t_rtttime = 0;
  579         /* Send one or 2 segments based on how much new data was acked. */
  580         if ((BYTES_THIS_ACK(tp, th) / tp->t_maxseg) >= 2)
  581                 num_segs = 2;
  582         tp->snd_cwnd = (tp->sackhint.sack_bytes_rexmit +
  583             (tp->snd_nxt - tp->sack_newdata) + num_segs * tp->t_maxseg);
  584         if (tp->snd_cwnd > tp->snd_ssthresh)
  585                 tp->snd_cwnd = tp->snd_ssthresh;
  586         tp->t_flags |= TF_ACKNOW;
  587         (void) tcp_output(tp);
  588 }
  589 
  590 #if 0
  591 /*
  592  * Debug version of tcp_sack_output() that walks the scoreboard.  Used for
  593  * now to sanity check the hint.
  594  */
  595 static struct sackhole *
  596 tcp_sack_output_debug(struct tcpcb *tp, int *sack_bytes_rexmt)
  597 {
  598         struct sackhole *p;
  599 
  600         INP_WLOCK_ASSERT(tp->t_inpcb);
  601         *sack_bytes_rexmt = 0;
  602         TAILQ_FOREACH(p, &tp->snd_holes, scblink) {
  603                 if (SEQ_LT(p->rxmit, p->end)) {
  604                         if (SEQ_LT(p->rxmit, tp->snd_una)) {/* old SACK hole */
  605                                 continue;
  606                         }
  607                         *sack_bytes_rexmt += (p->rxmit - p->start);
  608                         break;
  609                 }
  610                 *sack_bytes_rexmt += (p->rxmit - p->start);
  611         }
  612         return (p);
  613 }
  614 #endif
  615 
  616 /*
  617  * Returns the next hole to retransmit and the number of retransmitted bytes
  618  * from the scoreboard.  We store both the next hole and the number of
  619  * retransmitted bytes as hints (and recompute these on the fly upon SACK/ACK
  620  * reception).  This avoids scoreboard traversals completely.
  621  *
  622  * The loop here will traverse *at most* one link.  Here's the argument.  For
  623  * the loop to traverse more than 1 link before finding the next hole to
  624  * retransmit, we would need to have at least 1 node following the current
  625  * hint with (rxmit == end).  But, for all holes following the current hint,
  626  * (start == rxmit), since we have not yet retransmitted from them.
  627  * Therefore, in order to traverse more 1 link in the loop below, we need to
  628  * have at least one node following the current hint with (start == rxmit ==
  629  * end).  But that can't happen, (start == end) means that all the data in
  630  * that hole has been sacked, in which case, the hole would have been removed
  631  * from the scoreboard.
  632  */
  633 struct sackhole *
  634 tcp_sack_output(struct tcpcb *tp, int *sack_bytes_rexmt)
  635 {
  636         struct sackhole *hole = NULL;
  637 
  638         INP_WLOCK_ASSERT(tp->t_inpcb);
  639         *sack_bytes_rexmt = tp->sackhint.sack_bytes_rexmit;
  640         hole = tp->sackhint.nexthole;
  641         if (hole == NULL || SEQ_LT(hole->rxmit, hole->end))
  642                 goto out;
  643         while ((hole = TAILQ_NEXT(hole, scblink)) != NULL) {
  644                 if (SEQ_LT(hole->rxmit, hole->end)) {
  645                         tp->sackhint.nexthole = hole;
  646                         break;
  647                 }
  648         }
  649 out:
  650         return (hole);
  651 }
  652 
  653 /*
  654  * After a timeout, the SACK list may be rebuilt.  This SACK information
  655  * should be used to avoid retransmitting SACKed data.  This function
  656  * traverses the SACK list to see if snd_nxt should be moved forward.
  657  */
  658 void
  659 tcp_sack_adjust(struct tcpcb *tp)
  660 {
  661         struct sackhole *p, *cur = TAILQ_FIRST(&tp->snd_holes);
  662 
  663         INP_WLOCK_ASSERT(tp->t_inpcb);
  664         if (cur == NULL)
  665                 return; /* No holes */
  666         if (SEQ_GEQ(tp->snd_nxt, tp->snd_fack))
  667                 return; /* We're already beyond any SACKed blocks */
  668         /*-
  669          * Two cases for which we want to advance snd_nxt:
  670          * i) snd_nxt lies between end of one hole and beginning of another
  671          * ii) snd_nxt lies between end of last hole and snd_fack
  672          */
  673         while ((p = TAILQ_NEXT(cur, scblink)) != NULL) {
  674                 if (SEQ_LT(tp->snd_nxt, cur->end))
  675                         return;
  676                 if (SEQ_GEQ(tp->snd_nxt, p->start))
  677                         cur = p;
  678                 else {
  679                         tp->snd_nxt = p->start;
  680                         return;
  681                 }
  682         }
  683         if (SEQ_LT(tp->snd_nxt, cur->end))
  684                 return;
  685         tp->snd_nxt = tp->snd_fack;
  686 }

Cache object: da21ce4b17a1efbd732af1732361b990


[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]


This page is part of the FreeBSD/Linux Linux Kernel Cross-Reference, and was automatically generated using a modified version of the LXR engine.