-
Notifications
You must be signed in to change notification settings - Fork 144
/
Copy pathtcp_timer.c
1358 lines (1158 loc) · 38.7 KB
/
tcp_timer.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
* Copyright (c) 2000-2012 Apple Inc. All rights reserved.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
* This file contains Original Code and/or Modifications of Original Code
* as defined in and that are subject to the Apple Public Source License
* Version 2.0 (the 'License'). You may not use this file except in
* compliance with the License. The rights granted to you under the License
* may not be used to create, or enable the creation or redistribution of,
* unlawful or unlicensed copies of an Apple operating system, or to
* circumvent, violate, or enable the circumvention or violation of, any
* terms of an Apple operating system software license agreement.
*
* Please obtain a copy of the License at
* http://www.opensource.apple.com/apsl/ and read it before using this file.
*
* The Original Code and all software distributed under the License are
* distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
* Please see the License for the specific language governing rights and
* limitations under the License.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_END@
*/
/*
* Copyright (c) 1982, 1986, 1988, 1990, 1993, 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_timer.c 8.2 (Berkeley) 5/24/95
* $FreeBSD: src/sys/netinet/tcp_timer.c,v 1.34.2.11 2001/08/22 00:59:12 silby Exp $
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/mbuf.h>
#include <sys/sysctl.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/protosw.h>
#include <sys/domain.h>
#include <sys/mcache.h>
#include <sys/queue.h>
#include <kern/locks.h>
#include <kern/cpu_number.h> /* before tcp_seq.h, for tcp_random18() */
#include <net/route.h>
#include <net/if_var.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/in_pcb.h>
#if INET6
#include <netinet6/in6_pcb.h>
#endif
#include <netinet/ip_var.h>
#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/tcp_cc.h>
#if INET6
#include <netinet6/tcp6_var.h>
#endif
#include <netinet/tcpip.h>
#if TCPDEBUG
#include <netinet/tcp_debug.h>
#endif
#include <sys/kdebug.h>
#include <mach/sdt.h>
extern void postevent(struct socket *, struct sockbuf *,
int);
#define DBG_FNC_TCP_FAST NETDBG_CODE(DBG_NETTCP, (5 << 8))
#define DBG_FNC_TCP_SLOW NETDBG_CODE(DBG_NETTCP, (5 << 8) | 1)
#define TIMERENTRY_TO_TP(te) ((struct tcpcb *)((uintptr_t)te - offsetof(struct tcpcb, tentry.le.le_next)))
#define VERIFY_NEXT_LINK(elm,field) do { \
if (LIST_NEXT((elm),field) != NULL && \
LIST_NEXT((elm),field)->field.le_prev != \
&((elm)->field.le_next)) \
panic("Bad link elm %p next->prev != elm", (elm)); \
} while(0)
#define VERIFY_PREV_LINK(elm,field) do { \
if (*(elm)->field.le_prev != (elm)) \
panic("Bad link elm %p prev->next != elm", (elm)); \
} while(0)
static int background_io_trigger = 5;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, background_io_trigger, CTLFLAG_RW | CTLFLAG_LOCKED,
&background_io_trigger, 0, "Background IO Trigger Setting");
static int
sysctl_msec_to_ticks SYSCTL_HANDLER_ARGS
{
#pragma unused(arg1, arg2)
int error, s, tt;
tt = *(int *)oidp->oid_arg1;
s = tt * 1000 / TCP_RETRANSHZ;;
error = sysctl_handle_int(oidp, &s, 0, req);
if (error || !req->newptr)
return (error);
tt = s * TCP_RETRANSHZ / 1000;
if (tt < 1)
return (EINVAL);
*(int *)oidp->oid_arg1 = tt;
return (0);
}
int tcp_keepinit;
SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINIT, keepinit, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_keepinit, 0, sysctl_msec_to_ticks, "I", "");
int tcp_keepidle;
SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPIDLE, keepidle, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_keepidle, 0, sysctl_msec_to_ticks, "I", "");
int tcp_keepintvl;
SYSCTL_PROC(_net_inet_tcp, TCPCTL_KEEPINTVL, keepintvl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_keepintvl, 0, sysctl_msec_to_ticks, "I", "");
int tcp_msl;
SYSCTL_PROC(_net_inet_tcp, OID_AUTO, msl, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_msl, 0, sysctl_msec_to_ticks, "I", "Maximum segment lifetime");
/*
* Avoid DoS via TCP Robustness in Persist Condition (see http://www.ietf.org/id/draft-ananth-tcpm-persist-02.txt)
* by allowing a system wide maximum persistence timeout value when in Zero Window Probe mode.
* Expressed in milliseconds to be consistent without timeout related values, the TCP socket option is in seconds.
*/
u_int32_t tcp_max_persist_timeout = 0;
SYSCTL_PROC(_net_inet_tcp, OID_AUTO, max_persist_timeout, CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_max_persist_timeout, 0, sysctl_msec_to_ticks, "I", "Maximum persistence timout for ZWP");
static int always_keepalive = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, always_keepalive, CTLFLAG_RW | CTLFLAG_LOCKED,
&always_keepalive , 0, "Assume SO_KEEPALIVE on all TCP connections");
/* This parameter determines how long the timer list will stay in fast mode even
* though all connections are idle. In fast mode, the timer will fire more frequently
* anticipating new data.
*/
int timer_fastmode_idlemax = TCP_FASTMODE_IDLEGEN_MAX;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, timer_fastmode_idlemax, CTLFLAG_RW | CTLFLAG_LOCKED,
&timer_fastmode_idlemax, 0, "Maximum idle generations in fast mode");
/*
* See tcp_syn_backoff[] for interval values between SYN retransmits;
* the value set below defines the number of retransmits, before we
* disable the timestamp and window scaling options during subsequent
* SYN retransmits. Setting it to 0 disables the dropping off of those
* two options.
*/
static int tcp_broken_peer_syn_rxmit_thres = 7;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, broken_peer_syn_rxmit_thres, CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_broken_peer_syn_rxmit_thres, 0, "Number of retransmitted SYNs before "
"TCP disables rfc1323 and rfc1644 during the rest of attempts");
static int tcp_timer_advanced = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, tcp_timer_advanced, CTLFLAG_RD | CTLFLAG_LOCKED,
&tcp_timer_advanced, 0, "Number of times one of the timers was advanced");
static int tcp_resched_timerlist = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, tcp_resched_timerlist, CTLFLAG_RD | CTLFLAG_LOCKED,
&tcp_resched_timerlist, 0,
"Number of times timer list was rescheduled as part of processing a packet");
int tcp_pmtud_black_hole_detect = 1 ;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, pmtud_blackhole_detection, CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_pmtud_black_hole_detect, 0, "Path MTU Discovery Black Hole Detection");
int tcp_pmtud_black_hole_mss = 1200 ;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, pmtud_blackhole_mss, CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_pmtud_black_hole_mss, 0, "Path MTU Discovery Black Hole Detection lowered MSS");
static int tcp_keepcnt = TCPTV_KEEPCNT;
static int tcp_gc_done = FALSE; /* perfromed garbage collection of "used" sockets */
/* max idle probes */
int tcp_maxpersistidle;
/* max idle time in persist */
int tcp_maxidle;
/* TCP delack timer is set to 100 ms. Since the processing of timer list in fast
* mode will happen no faster than 100 ms, the delayed ack timer will fire some where
* between 100 and 200 ms.
*/
int tcp_delack = TCP_RETRANSHZ / 10;
struct inpcbhead time_wait_slots[N_TIME_WAIT_SLOTS];
int cur_tw_slot = 0;
/* tcp timer list */
struct tcptimerlist tcp_timer_list;
/* The frequency of running through the TCP timer list in
* fast and slow mode can be configured.
*/
SYSCTL_UINT(_net_inet_tcp, OID_AUTO, timer_fastquantum, CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_timer_list.fast_quantum, TCP_FASTTIMER_QUANTUM,
"Frequency of running timer list in fast mode");
SYSCTL_UINT(_net_inet_tcp, OID_AUTO, timer_slowquantum, CTLFLAG_RW | CTLFLAG_LOCKED,
&tcp_timer_list.slow_quantum, TCP_SLOWTIMER_QUANTUM,
"Frequency of running timer list in slow mode");
static void tcp_remove_timer(struct tcpcb *tp);
static void tcp_sched_timerlist(uint32_t offset);
static uint32_t tcp_run_conn_timer(struct tcpcb *tp, uint16_t *next_index);
static void tcp_sched_timers(struct tcpcb *tp);
static inline void tcp_set_lotimer_index(struct tcpcb *);
/* Macro to compare two timers. If there is a reset of the sign bit, it is
* safe to assume that the timer has wrapped around. By doing signed comparision,
* we take care of wrap around such that the value with the sign bit reset is
* actually ahead of the other.
*/
static inline int32_t
timer_diff(uint32_t t1, uint32_t toff1, uint32_t t2, uint32_t toff2) {
return (int32_t)((t1 + toff1) - (t2 + toff2));
};
/* Returns true if the timer is on the timer list */
#define TIMER_IS_ON_LIST(tp) ((tp)->t_flags & TF_TIMER_ONLIST)
void add_to_time_wait_locked(struct tcpcb *tp, uint32_t delay);
void add_to_time_wait(struct tcpcb *tp, uint32_t delay) ;
static void tcp_garbage_collect(struct inpcb *, int);
void add_to_time_wait_locked(struct tcpcb *tp, uint32_t delay)
{
int tw_slot;
struct inpcbinfo *pcbinfo = &tcbinfo;
uint32_t timer;
/* pcb list should be locked when we get here */
lck_rw_assert(pcbinfo->mtx, LCK_RW_ASSERT_EXCLUSIVE);
LIST_REMOVE(tp->t_inpcb, inp_list);
/* if (tp->t_timer[TCPT_2MSL] <= 0)
tp->t_timer[TCPT_2MSL] = 1; */
/*
* Because we're pulling this pcb out of the main TCP pcb list,
* we need to recalculate the TCPT_2MSL timer value for tcp_slowtimo
* higher timer granularity.
*/
timer = (delay / TCP_RETRANSHZ) * PR_SLOWHZ;
tp->t_rcvtime = (tp->t_rcvtime / TCP_RETRANSHZ) * PR_SLOWHZ;
tp->t_rcvtime += timer & (N_TIME_WAIT_SLOTS - 1);
tw_slot = (timer & (N_TIME_WAIT_SLOTS - 1)) + cur_tw_slot;
if (tw_slot >= N_TIME_WAIT_SLOTS)
tw_slot -= N_TIME_WAIT_SLOTS;
LIST_INSERT_HEAD(&time_wait_slots[tw_slot], tp->t_inpcb, inp_list);
}
void add_to_time_wait(struct tcpcb *tp, uint32_t delay)
{
struct inpcbinfo *pcbinfo = &tcbinfo;
if (!lck_rw_try_lock_exclusive(pcbinfo->mtx)) {
tcp_unlock(tp->t_inpcb->inp_socket, 0, 0);
lck_rw_lock_exclusive(pcbinfo->mtx);
tcp_lock(tp->t_inpcb->inp_socket, 0, 0);
}
add_to_time_wait_locked(tp, delay);
lck_rw_done(pcbinfo->mtx);
}
static void
tcp_garbage_collect(struct inpcb *inp, int istimewait)
{
struct socket *so;
struct tcpcb *tp;
so = inp->inp_socket;
tp = intotcpcb(inp);
/*
* Skip if still in use or busy; it would have been more efficient
* if we were to test so_usecount against 0, but this isn't possible
* due to the current implementation of tcp_dropdropablreq() where
* overflow sockets that are eligible for garbage collection have
* their usecounts set to 1.
*/
if (so->so_usecount > 1 || !lck_mtx_try_lock_spin(&inp->inpcb_mtx))
return;
/* Check again under the lock */
if (so->so_usecount > 1) {
lck_mtx_unlock(&inp->inpcb_mtx);
return;
}
/*
* Overflowed socket dropped from the listening queue? Do this
* only if we are called to clean up the time wait slots, since
* tcp_dropdropablreq() considers a socket to have been fully
* dropped after add_to_time_wait() is finished.
* Also handle the case of connections getting closed by the peer while in the queue as
* seen with rdar://6422317
*
*/
if (so->so_usecount == 1 &&
((istimewait && (so->so_flags & SOF_OVERFLOW)) ||
((tp != NULL) && (tp->t_state == TCPS_CLOSED) && (so->so_head != NULL)
&& ((so->so_state & (SS_INCOMP|SS_CANTSENDMORE|SS_CANTRCVMORE)) ==
(SS_INCOMP|SS_CANTSENDMORE|SS_CANTRCVMORE))))) {
if (inp->inp_state != INPCB_STATE_DEAD) {
/* Become a regular mutex */
lck_mtx_convert_spin(&inp->inpcb_mtx);
#if INET6
if (INP_CHECK_SOCKAF(so, AF_INET6))
in6_pcbdetach(inp);
else
#endif /* INET6 */
in_pcbdetach(inp);
}
so->so_usecount--;
lck_mtx_unlock(&inp->inpcb_mtx);
return;
} else if (inp->inp_wantcnt != WNT_STOPUSING) {
lck_mtx_unlock(&inp->inpcb_mtx);
return;
}
/*
* We get here because the PCB is no longer searchable (WNT_STOPUSING);
* detach (if needed) and dispose if it is dead (usecount is 0). This
* covers all cases, including overflow sockets and those that are
* considered as "embryonic", i.e. created by sonewconn() in TCP input
* path, and have not yet been committed. For the former, we reduce
* the usecount to 0 as done by the code above. For the latter, the
* usecount would have reduced to 0 as part calling soabort() when the
* socket is dropped at the end of tcp_input().
*/
if (so->so_usecount == 0) {
DTRACE_TCP4(state__change, void, NULL, struct inpcb *, inp,
struct tcpcb *, tp, int32_t, TCPS_CLOSED);
/* Become a regular mutex */
lck_mtx_convert_spin(&inp->inpcb_mtx);
/* If this tp still happens to be on the timer list,
* take it out
*/
if (TIMER_IS_ON_LIST(tp)) {
tcp_remove_timer(tp);
}
if (inp->inp_state != INPCB_STATE_DEAD) {
#if INET6
if (INP_CHECK_SOCKAF(so, AF_INET6))
in6_pcbdetach(inp);
else
#endif /* INET6 */
in_pcbdetach(inp);
}
in_pcbdispose(inp);
} else {
lck_mtx_unlock(&inp->inpcb_mtx);
}
}
void
tcp_slowtimo(void)
{
struct inpcb *inp, *nxt;
struct tcpcb *tp;
#if TCPDEBUG
int ostate;
#endif
#if KDEBUG
static int tws_checked = 0;
#endif
struct inpcbinfo *pcbinfo = &tcbinfo;
KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_START, 0,0,0,0,0);
tcp_maxidle = tcp_keepcnt * tcp_keepintvl;
/* Update tcp_now here as it may get used while processing the slow timer */
calculate_tcp_clock();
/* Garbage collect socket/tcpcb: We need to acquire the list lock
* exclusively to do this
*/
if (lck_rw_try_lock_exclusive(pcbinfo->mtx) == FALSE) {
if (tcp_gc_done == TRUE) { /* don't sweat it this time. cleanup was done last time */
tcp_gc_done = FALSE;
KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_END, tws_checked, cur_tw_slot,0,0,0);
return; /* Upgrade failed and lost lock - give up this time. */
}
lck_rw_lock_exclusive(pcbinfo->mtx); /* Upgrade failed, lost lock now take it again exclusive */
}
tcp_gc_done = TRUE;
/*
* Process the items in the current time-wait slot
*/
#if KDEBUG
tws_checked = 0;
#endif
KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_NONE, tws_checked,0,0,0,0);
LIST_FOREACH(inp, &time_wait_slots[cur_tw_slot], inp_list) {
#if KDEBUG
tws_checked++;
#endif
if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING)
continue;
tcp_lock(inp->inp_socket, 1, 0);
if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING)
goto twunlock;
tp = intotcpcb(inp);
if (tp == NULL) /* tp already closed, remove from list */
goto twunlock;
if (tp->t_timer[TCPT_2MSL] >= N_TIME_WAIT_SLOTS) {
tp->t_timer[TCPT_2MSL] -= N_TIME_WAIT_SLOTS;
tp->t_rcvtime += N_TIME_WAIT_SLOTS;
}
else
tp->t_timer[TCPT_2MSL] = 0;
if (tp->t_timer[TCPT_2MSL] == 0) {
/* That pcb is ready for a close */
tcp_free_sackholes(tp);
tp = tcp_close(tp);
}
twunlock:
tcp_unlock(inp->inp_socket, 1, 0);
}
LIST_FOREACH_SAFE(inp, &tcb, inp_list, nxt) {
tcp_garbage_collect(inp, 0);
}
/* Now cleanup the time wait ones */
LIST_FOREACH_SAFE(inp, &time_wait_slots[cur_tw_slot], inp_list, nxt) {
tcp_garbage_collect(inp, 1);
}
if (++cur_tw_slot >= N_TIME_WAIT_SLOTS)
cur_tw_slot = 0;
lck_rw_done(pcbinfo->mtx);
KERNEL_DEBUG(DBG_FNC_TCP_SLOW | DBG_FUNC_END, tws_checked, cur_tw_slot,0,0,0);
}
/*
* Cancel all timers for TCP tp.
*/
void
tcp_canceltimers(tp)
struct tcpcb *tp;
{
register int i;
tcp_remove_timer(tp);
for (i = 0; i < TCPT_NTIMERS; i++)
tp->t_timer[i] = 0;
tp->tentry.timer_start = tcp_now;
tp->tentry.index = TCPT_NONE;
}
int tcp_syn_backoff[TCP_MAXRXTSHIFT + 1] =
{ 1, 1, 1, 1, 1, 2, 4, 8, 16, 32, 64, 64, 64 };
int tcp_backoff[TCP_MAXRXTSHIFT + 1] =
{ 1, 2, 4, 8, 16, 32, 64, 64, 64, 64, 64, 64, 64 };
static int tcp_totbackoff = 511; /* sum of tcp_backoff[] */
/*
* TCP timer processing.
*/
struct tcpcb *
tcp_timers(tp, timer)
register struct tcpcb *tp;
int timer;
{
register int rexmt;
struct socket *so;
struct tcptemp *t_template;
int optlen = 0;
int idle_time = 0;
#if TCPDEBUG
int ostate;
#endif
#if INET6
int isipv6 = (tp->t_inpcb->inp_vflag & INP_IPV4) == 0;
#endif /* INET6 */
so = tp->t_inpcb->inp_socket;
idle_time = tcp_now - tp->t_rcvtime;
switch (timer) {
/*
* 2 MSL timeout in shutdown went off. If we're closed but
* still waiting for peer to close and connection has been idle
* too long, or if 2MSL time is up from TIME_WAIT or FIN_WAIT_2,
* delete connection control block.
* Otherwise, (this case shouldn't happen) check again in a bit
* we keep the socket in the main list in that case.
*/
case TCPT_2MSL:
tcp_free_sackholes(tp);
if (tp->t_state != TCPS_TIME_WAIT &&
tp->t_state != TCPS_FIN_WAIT_2 &&
((idle_time > 0) && (idle_time < tcp_maxidle))) {
tp->t_timer[TCPT_2MSL] = OFFSET_FROM_START(tp, (u_int32_t)tcp_keepintvl);
}
else {
tp = tcp_close(tp);
return(tp);
}
break;
/*
* Retransmission timer went off. Message has not
* been acked within retransmit interval. Back off
* to a longer retransmit interval and retransmit one segment.
*/
case TCPT_REXMT:
/* Drop a connection in the retransmit timer
* 1. If we have retransmitted more than TCP_MAXRXTSHIFT times
* 2. If the time spent in this retransmission episode is more than
* the time limit set with TCP_RXT_CONNDROPTIME socket option
* 3. If TCP_RXT_FINDROP socket option was set and we have already
* retransmitted the FIN 3 times without receiving an ack
*/
if (++tp->t_rxtshift > TCP_MAXRXTSHIFT ||
(tp->rxt_conndroptime > 0 && tp->rxt_start > 0 &&
(tcp_now - tp->rxt_start) >= tp->rxt_conndroptime) ||
((tp->t_flagsext & TF_RXTFINDROP) != 0 &&
(tp->t_flags & TF_SENTFIN) != 0 &&
tp->t_rxtshift >= 4)) {
if ((tp->t_flagsext & TF_RXTFINDROP) != 0) {
tcpstat.tcps_rxtfindrop++;
} else {
tcpstat.tcps_timeoutdrop++;
}
tp->t_rxtshift = TCP_MAXRXTSHIFT;
postevent(so, 0, EV_TIMEOUT);
soevent(so,
(SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT));
tp = tcp_drop(tp, tp->t_softerror ?
tp->t_softerror : ETIMEDOUT);
break;
}
if (tp->t_rxtshift == 1) {
/*
* first retransmit; record ssthresh and cwnd so they can
* be recovered if this turns out to be a "bad" retransmit.
* A retransmit is considered "bad" if an ACK for this
* segment is received within RTT/2 interval; the assumption
* here is that the ACK was already in flight. See
* "On Estimating End-to-End Network Path Properties" by
* Allman and Paxson for more details.
*/
tp->snd_cwnd_prev = tp->snd_cwnd;
tp->snd_ssthresh_prev = tp->snd_ssthresh;
tp->snd_recover_prev = tp->snd_recover;
if (IN_FASTRECOVERY(tp))
tp->t_flags |= TF_WASFRECOVERY;
else
tp->t_flags &= ~TF_WASFRECOVERY;
tp->t_badrxtwin = tcp_now + (tp->t_srtt >> (TCP_RTT_SHIFT));
/* Set the time at which retransmission on this
* connection started
*/
tp->rxt_start = tcp_now;
}
tcpstat.tcps_rexmttimeo++;
if (tp->t_state == TCPS_SYN_SENT)
rexmt = TCP_REXMTVAL(tp) * tcp_syn_backoff[tp->t_rxtshift];
else
rexmt = TCP_REXMTVAL(tp) * tcp_backoff[tp->t_rxtshift];
TCPT_RANGESET(tp->t_rxtcur, rexmt,
tp->t_rttmin, TCPTV_REXMTMAX,
TCP_ADD_REXMTSLOP(tp));
tp->t_timer[TCPT_REXMT] = OFFSET_FROM_START(tp, tp->t_rxtcur);
if (INP_WAIT_FOR_IF_FEEDBACK(tp->t_inpcb))
goto fc_output;
tcp_free_sackholes(tp);
/*
* Check for potential Path MTU Discovery Black Hole
*/
if (tcp_pmtud_black_hole_detect && (tp->t_state == TCPS_ESTABLISHED)) {
if (((tp->t_flags & (TF_PMTUD|TF_MAXSEGSNT)) == (TF_PMTUD|TF_MAXSEGSNT)) &&
(tp->t_rxtshift == 2)) {
/*
* Enter Path MTU Black-hole Detection mechanism:
* - Disable Path MTU Discovery (IP "DF" bit).
* - Reduce MTU to lower value than what we negociated with peer.
*/
tp->t_flags &= ~TF_PMTUD; /* Disable Path MTU Discovery for now */
tp->t_flags |= TF_BLACKHOLE; /* Record that we may have found a black hole */
optlen = tp->t_maxopd - tp->t_maxseg;
tp->t_pmtud_saved_maxopd = tp->t_maxopd; /* Keep track of previous MSS */
if (tp->t_maxopd > tcp_pmtud_black_hole_mss)
tp->t_maxopd = tcp_pmtud_black_hole_mss; /* Reduce the MSS to intermediary value */
else {
tp->t_maxopd = /* use the default MSS */
#if INET6
isipv6 ? tcp_v6mssdflt :
#endif /* INET6 */
tcp_mssdflt;
}
tp->t_maxseg = tp->t_maxopd - optlen;
/*
* Reset the slow-start flight size as it may depends on the new MSS
*/
if (CC_ALGO(tp)->cwnd_init != NULL)
CC_ALGO(tp)->cwnd_init(tp);
}
/*
* If further retransmissions are still unsuccessful with a lowered MTU,
* maybe this isn't a Black Hole and we restore the previous MSS and
* blackhole detection flags.
*/
else {
if ((tp->t_flags & TF_BLACKHOLE) && (tp->t_rxtshift > 4)) {
tp->t_flags |= TF_PMTUD;
tp->t_flags &= ~TF_BLACKHOLE;
optlen = tp->t_maxopd - tp->t_maxseg;
tp->t_maxopd = tp->t_pmtud_saved_maxopd;
tp->t_maxseg = tp->t_maxopd - optlen;
/*
* Reset the slow-start flight size as it may depends on the new MSS
*/
if (CC_ALGO(tp)->cwnd_init != NULL)
CC_ALGO(tp)->cwnd_init(tp);
}
}
}
/*
* Disable rfc1323 and rfc1644 if we haven't got any response to
* our SYN (after we reach the threshold) to work-around some
* broken terminal servers (most of which have hopefully been
* retired) that have bad VJ header compression code which
* trashes TCP segments containing unknown-to-them TCP options.
*/
if ((tp->t_state == TCPS_SYN_SENT) &&
(tp->t_rxtshift == tcp_broken_peer_syn_rxmit_thres))
tp->t_flags &= ~(TF_REQ_SCALE|TF_REQ_TSTMP|TF_REQ_CC);
/*
* If losing, let the lower level know and try for
* a better route. Also, if we backed off this far,
* our srtt estimate is probably bogus. Clobber it
* so we'll take the next rtt measurement as our srtt;
* move the current srtt into rttvar to keep the current
* retransmit times until then.
*/
if (tp->t_rxtshift > TCP_MAXRXTSHIFT / 4) {
#if INET6
if (isipv6)
in6_losing(tp->t_inpcb);
else
#endif /* INET6 */
in_losing(tp->t_inpcb);
tp->t_rttvar += (tp->t_srtt >> TCP_RTT_SHIFT);
tp->t_srtt = 0;
}
tp->snd_nxt = tp->snd_una;
/*
* Note: We overload snd_recover to function also as the
* snd_last variable described in RFC 2582
*/
tp->snd_recover = tp->snd_max;
/*
* Force a segment to be sent.
*/
tp->t_flags |= TF_ACKNOW;
/*
* If timing a segment in this window, stop the timer.
*/
tp->t_rtttime = 0;
if (CC_ALGO(tp)->after_timeout != NULL)
CC_ALGO(tp)->after_timeout(tp);
tp->t_dupacks = 0;
EXIT_FASTRECOVERY(tp);
/* CWR notifications are to be sent on new data right after
* RTOs, Fast Retransmits and ECE notification receipts.
*/
if ((tp->ecn_flags & TE_ECN_ON) == TE_ECN_ON) {
tp->ecn_flags |= TE_SENDCWR;
}
fc_output:
DTRACE_TCP5(cc, void, NULL, struct inpcb *, tp->t_inpcb,
struct tcpcb *, tp, struct tcphdr *, NULL,
int32_t, TCP_CC_REXMT_TIMEOUT);
(void) tcp_output(tp);
break;
/*
* Persistance timer into zero window.
* Force a byte to be output, if possible.
*/
case TCPT_PERSIST:
tcpstat.tcps_persisttimeo++;
/*
* Hack: if the peer is dead/unreachable, we do not
* time out if the window is closed. After a full
* backoff, drop the connection if the idle time
* (no responses to probes) reaches the maximum
* backoff that we would use if retransmitting.
*
* Drop the connection if we reached the maximum allowed time for
* Zero Window Probes without a non-zero update from the peer.
* See rdar://5805356
*/
if ((tp->t_rxtshift == TCP_MAXRXTSHIFT &&
(idle_time >= tcp_maxpersistidle ||
idle_time >= TCP_REXMTVAL(tp) * tcp_totbackoff)) ||
((tp->t_persist_stop != 0) &&
TSTMP_LEQ(tp->t_persist_stop, tcp_now))) {
tcpstat.tcps_persistdrop++;
postevent(so, 0, EV_TIMEOUT);
soevent(so,
(SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT));
tp = tcp_drop(tp, ETIMEDOUT);
break;
}
tcp_setpersist(tp);
tp->t_force = 1;
(void) tcp_output(tp);
tp->t_force = 0;
break;
/*
* Keep-alive timer went off; send something
* or drop connection if idle for too long.
*/
case TCPT_KEEP:
tcpstat.tcps_keeptimeo++;
if (tp->t_state < TCPS_ESTABLISHED)
goto dropit;
if ((always_keepalive ||
tp->t_inpcb->inp_socket->so_options & SO_KEEPALIVE) &&
(tp->t_state <= TCPS_CLOSING || tp->t_state == TCPS_FIN_WAIT_2)) {
if (idle_time >= TCP_KEEPIDLE(tp) + (u_int32_t)tcp_maxidle)
goto dropit;
/*
* Send a packet designed to force a response
* if the peer is up and reachable:
* either an ACK if the connection is still alive,
* or an RST if the peer has closed the connection
* due to timeout or reboot.
* Using sequence number tp->snd_una-1
* causes the transmitted zero-length segment
* to lie outside the receive window;
* by the protocol spec, this requires the
* correspondent TCP to respond.
*/
tcpstat.tcps_keepprobe++;
t_template = tcp_maketemplate(tp);
if (t_template) {
unsigned int ifscope, nocell = 0;
if (tp->t_inpcb->inp_flags & INP_BOUND_IF)
ifscope = tp->t_inpcb->inp_boundifp->if_index;
else
ifscope = IFSCOPE_NONE;
/*
* If the socket isn't allowed to use the
* cellular interface, indicate it as such.
*/
if (tp->t_inpcb->inp_flags & INP_NO_IFT_CELLULAR)
nocell = 1;
tcp_respond(tp, t_template->tt_ipgen,
&t_template->tt_t, (struct mbuf *)NULL,
tp->rcv_nxt, tp->snd_una - 1, 0, ifscope,
nocell);
(void) m_free(dtom(t_template));
}
tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START(tp, tcp_keepintvl);
} else
tp->t_timer[TCPT_KEEP] = OFFSET_FROM_START(tp, TCP_KEEPIDLE(tp));
break;
case TCPT_DELACK:
if (tcp_delack_enabled && (tp->t_flags & TF_DELACK)) {
tp->t_flags &= ~TF_DELACK;
tp->t_timer[TCPT_DELACK] = 0;
tp->t_flags |= TF_ACKNOW;
/* If delayed ack timer fired while we are stretching acks,
* go back to acking every other packet
*/
if ((tp->t_flags & TF_STRETCHACK) != 0)
tcp_reset_stretch_ack(tp);
/* If we are measuring inter packet arrival jitter for
* throttling a connection, this delayed ack might be
* the reason for accumulating some jitter. So let's
* restart the measurement.
*/
CLEAR_IAJ_STATE(tp);
tcpstat.tcps_delack++;
(void) tcp_output(tp);
}
break;
#if TCPDEBUG
if (tp->t_inpcb->inp_socket->so_options & SO_DEBUG)
tcp_trace(TA_USER, ostate, tp, (void *)0, (struct tcphdr *)0,
PRU_SLOWTIMO);
#endif
dropit:
tcpstat.tcps_keepdrops++;
postevent(so, 0, EV_TIMEOUT);
soevent(so,
(SO_FILT_HINT_LOCKED|SO_FILT_HINT_TIMEOUT));
tp = tcp_drop(tp, ETIMEDOUT);
break;
}
return (tp);
}
/* Remove a timer entry from timer list */
void
tcp_remove_timer(struct tcpcb *tp)
{
struct tcptimerlist *listp = &tcp_timer_list;
lck_mtx_assert(&tp->t_inpcb->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
if (!(TIMER_IS_ON_LIST(tp))) {
return;
}
lck_mtx_lock(listp->mtx);
/* Check if pcb is on timer list again after acquiring the lock */
if (!(TIMER_IS_ON_LIST(tp))) {
lck_mtx_unlock(listp->mtx);
return;
}
if (listp->next_te != NULL && listp->next_te == &tp->tentry)
listp->next_te = LIST_NEXT(&tp->tentry, le);
LIST_REMOVE(&tp->tentry, le);
tp->t_flags &= ~(TF_TIMER_ONLIST);
listp->entries--;
tp->tentry.le.le_next = NULL;
tp->tentry.le.le_prev = NULL;
lck_mtx_unlock(listp->mtx);
}
/* Function to check if the timerlist needs to be rescheduled to run
* the timer entry correctly. Basically, this is to check if we can avoid
* taking the list lock.
*/
static boolean_t
need_to_resched_timerlist(uint32_t runtime, uint16_t index) {
struct tcptimerlist *listp = &tcp_timer_list;
int32_t diff;
boolean_t is_fast;
if (runtime == 0 || index == TCPT_NONE)
return FALSE;
is_fast = !(IS_TIMER_SLOW(index));
/* If the list is being processed then the state of the list is in flux.
* In this case always acquire the lock and set the state correctly.
*/
if (listp->running) {
return TRUE;
}
diff = timer_diff(listp->runtime, 0, runtime, 0);
if (diff <= 0) {
/* The list is going to run before this timer */
return FALSE;
} else {
if (is_fast) {
if (diff <= listp->fast_quantum)
return FALSE;
} else {
if (diff <= listp->slow_quantum)
return FALSE;
}
}
return TRUE;
}
void
tcp_sched_timerlist(uint32_t offset)
{
uint64_t deadline = 0;
struct tcptimerlist *listp = &tcp_timer_list;
lck_mtx_assert(listp->mtx, LCK_MTX_ASSERT_OWNED);
listp->runtime = tcp_now + offset;
clock_interval_to_deadline(offset, NSEC_PER_SEC / TCP_RETRANSHZ,
&deadline);
thread_call_enter_delayed(listp->call, deadline);
}
/* Function to run the timers for a connection.
*
* Returns the offset of next timer to be run for this connection which
* can be used to reschedule the timerlist.
*/
uint32_t
tcp_run_conn_timer(struct tcpcb *tp, uint16_t *next_index) {
struct socket *so;
uint16_t i = 0, index = TCPT_NONE, lo_index = TCPT_NONE;
uint32_t timer_val, offset = 0, lo_timer = 0;
int32_t diff;
boolean_t needtorun[TCPT_NTIMERS];
int count = 0;