summaryrefslogtreecommitdiffstats
path: root/sys/kern/kern_ktrace.c
blob: d4b722a91725ed4e750c5d0106cd400801d8f297 (plain)
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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
/*-
 * Copyright (c) 1989, 1993
 *	The Regents of the University of California.
 * Copyright (c) 2005 Robert N. M. Watson
 * 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.
 * 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.
 *
 *	@(#)kern_ktrace.c	8.2 (Berkeley) 9/23/93
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include "opt_ktrace.h"

#include <sys/param.h>
#include <sys/capability.h>
#include <sys/systm.h>
#include <sys/fcntl.h>
#include <sys/kernel.h>
#include <sys/kthread.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/malloc.h>
#include <sys/mount.h>
#include <sys/namei.h>
#include <sys/priv.h>
#include <sys/proc.h>
#include <sys/unistd.h>
#include <sys/vnode.h>
#include <sys/socket.h>
#include <sys/stat.h>
#include <sys/ktrace.h>
#include <sys/sx.h>
#include <sys/sysctl.h>
#include <sys/sysent.h>
#include <sys/syslog.h>
#include <sys/sysproto.h>

#include <security/mac/mac_framework.h>

/*
 * The ktrace facility allows the tracing of certain key events in user space
 * processes, such as system calls, signal delivery, context switches, and
 * user generated events using utrace(2).  It works by streaming event
 * records and data to a vnode associated with the process using the
 * ktrace(2) system call.  In general, records can be written directly from
 * the context that generates the event.  One important exception to this is
 * during a context switch, where sleeping is not permitted.  To handle this
 * case, trace events are generated using in-kernel ktr_request records, and
 * then delivered to disk at a convenient moment -- either immediately, the
 * next traceable event, at system call return, or at process exit.
 *
 * When dealing with multiple threads or processes writing to the same event
 * log, ordering guarantees are weak: specifically, if an event has multiple
 * records (i.e., system call enter and return), they may be interlaced with
 * records from another event.  Process and thread ID information is provided
 * in the record, and user applications can de-interlace events if required.
 */

static MALLOC_DEFINE(M_KTRACE, "KTRACE", "KTRACE");

#ifdef KTRACE

FEATURE(ktrace, "Kernel support for system-call tracing");

#ifndef KTRACE_REQUEST_POOL
#define	KTRACE_REQUEST_POOL	100
#endif

struct ktr_request {
	struct	ktr_header ktr_header;
	void	*ktr_buffer;
	union {
		struct	ktr_proc_ctor ktr_proc_ctor;
		struct	ktr_cap_fail ktr_cap_fail;
		struct	ktr_syscall ktr_syscall;
		struct	ktr_sysret ktr_sysret;
		struct	ktr_genio ktr_genio;
		struct	ktr_psig ktr_psig;
		struct	ktr_csw ktr_csw;
		struct	ktr_fault ktr_fault;
		struct	ktr_faultend ktr_faultend;
	} ktr_data;
	STAILQ_ENTRY(ktr_request) ktr_list;
};

static int data_lengths[] = {
	0,					/* none */
	offsetof(struct ktr_syscall, ktr_args),	/* KTR_SYSCALL */
	sizeof(struct ktr_sysret),		/* KTR_SYSRET */
	0,					/* KTR_NAMEI */
	sizeof(struct ktr_genio),		/* KTR_GENIO */
	sizeof(struct ktr_psig),		/* KTR_PSIG */
	sizeof(struct ktr_csw),			/* KTR_CSW */
	0,					/* KTR_USER */
	0,					/* KTR_STRUCT */
	0,					/* KTR_SYSCTL */
	sizeof(struct ktr_proc_ctor),		/* KTR_PROCCTOR */
	0,					/* KTR_PROCDTOR */
	sizeof(struct ktr_cap_fail),		/* KTR_CAPFAIL */
	sizeof(struct ktr_fault),		/* KTR_FAULT */
	sizeof(struct ktr_faultend),		/* KTR_FAULTEND */
};

static STAILQ_HEAD(, ktr_request) ktr_free;

static SYSCTL_NODE(_kern, OID_AUTO, ktrace, CTLFLAG_RD, 0, "KTRACE options");

static u_int ktr_requestpool = KTRACE_REQUEST_POOL;
TUNABLE_INT("kern.ktrace.request_pool", &ktr_requestpool);

static u_int ktr_geniosize = PAGE_SIZE;
TUNABLE_INT("kern.ktrace.genio_size", &ktr_geniosize);
SYSCTL_UINT(_kern_ktrace, OID_AUTO, genio_size, CTLFLAG_RW, &ktr_geniosize,
    0, "Maximum size of genio event payload");

static int print_message = 1;
static struct mtx ktrace_mtx;
static struct sx ktrace_sx;

static void ktrace_init(void *dummy);
static int sysctl_kern_ktrace_request_pool(SYSCTL_HANDLER_ARGS);
static u_int ktrace_resize_pool(u_int oldsize, u_int newsize);
static struct ktr_request *ktr_getrequest_entered(struct thread *td, int type);
static struct ktr_request *ktr_getrequest(int type);
static void ktr_submitrequest(struct thread *td, struct ktr_request *req);
static void ktr_freeproc(struct proc *p, struct ucred **uc,
    struct vnode **vp);
static void ktr_freerequest(struct ktr_request *req);
static void ktr_freerequest_locked(struct ktr_request *req);
static void ktr_writerequest(struct thread *td, struct ktr_request *req);
static int ktrcanset(struct thread *,struct proc *);
static int ktrsetchildren(struct thread *,struct proc *,int,int,struct vnode *);
static int ktrops(struct thread *,struct proc *,int,int,struct vnode *);
static void ktrprocctor_entered(struct thread *, struct proc *);

/*
 * ktrace itself generates events, such as context switches, which we do not
 * wish to trace.  Maintain a flag, TDP_INKTRACE, on each thread to determine
 * whether or not it is in a region where tracing of events should be
 * suppressed.
 */
static void
ktrace_enter(struct thread *td)
{

	KASSERT(!(td->td_pflags & TDP_INKTRACE), ("ktrace_enter: flag set"));
	td->td_pflags |= TDP_INKTRACE;
}

static void
ktrace_exit(struct thread *td)
{

	KASSERT(td->td_pflags & TDP_INKTRACE, ("ktrace_exit: flag not set"));
	td->td_pflags &= ~TDP_INKTRACE;
}

static void
ktrace_assert(struct thread *td)
{

	KASSERT(td->td_pflags & TDP_INKTRACE, ("ktrace_assert: flag not set"));
}

static void
ktrace_init(void *dummy)
{
	struct ktr_request *req;
	int i;

	mtx_init(&ktrace_mtx, "ktrace", NULL, MTX_DEF | MTX_QUIET);
	sx_init(&ktrace_sx, "ktrace_sx");
	STAILQ_INIT(&ktr_free);
	for (i = 0; i < ktr_requestpool; i++) {
		req = malloc(sizeof(struct ktr_request), M_KTRACE, M_WAITOK);
		STAILQ_INSERT_HEAD(&ktr_free, req, ktr_list);
	}
}
SYSINIT(ktrace_init, SI_SUB_KTRACE, SI_ORDER_ANY, ktrace_init, NULL);

static int
sysctl_kern_ktrace_request_pool(SYSCTL_HANDLER_ARGS)
{
	struct thread *td;
	u_int newsize, oldsize, wantsize;
	int error;

	/* Handle easy read-only case first to avoid warnings from GCC. */
	if (!req->newptr) {
		oldsize = ktr_requestpool;
		return (SYSCTL_OUT(req, &oldsize, sizeof(u_int)));
	}

	error = SYSCTL_IN(req, &wantsize, sizeof(u_int));
	if (error)
		return (error);
	td = curthread;
	ktrace_enter(td);
	oldsize = ktr_requestpool;
	newsize = ktrace_resize_pool(oldsize, wantsize);
	ktrace_exit(td);
	error = SYSCTL_OUT(req, &oldsize, sizeof(u_int));
	if (error)
		return (error);
	if (wantsize > oldsize && newsize < wantsize)
		return (ENOSPC);
	return (0);
}
SYSCTL_PROC(_kern_ktrace, OID_AUTO, request_pool, CTLTYPE_UINT|CTLFLAG_RW,
    &ktr_requestpool, 0, sysctl_kern_ktrace_request_pool, "IU",
    "Pool buffer size for ktrace(1)");

static u_int
ktrace_resize_pool(u_int oldsize, u_int newsize)
{
	STAILQ_HEAD(, ktr_request) ktr_new;
	struct ktr_request *req;
	int bound;

	print_message = 1;
	bound = newsize - oldsize;
	if (bound == 0)
		return (ktr_requestpool);
	if (bound < 0) {
		mtx_lock(&ktrace_mtx);
		/* Shrink pool down to newsize if possible. */
		while (bound++ < 0) {
			req = STAILQ_FIRST(&ktr_free);
			if (req == NULL)
				break;
			STAILQ_REMOVE_HEAD(&ktr_free, ktr_list);
			ktr_requestpool--;
			free(req, M_KTRACE);
		}
	} else {
		/* Grow pool up to newsize. */
		STAILQ_INIT(&ktr_new);
		while (bound-- > 0) {
			req = malloc(sizeof(struct ktr_request), M_KTRACE,
			    M_WAITOK);
			STAILQ_INSERT_HEAD(&ktr_new, req, ktr_list);
		}
		mtx_lock(&ktrace_mtx);
		STAILQ_CONCAT(&ktr_free, &ktr_new);
		ktr_requestpool += (newsize - oldsize);
	}
	mtx_unlock(&ktrace_mtx);
	return (ktr_requestpool);
}

/* ktr_getrequest() assumes that ktr_comm[] is the same size as td_name[]. */
CTASSERT(sizeof(((struct ktr_header *)NULL)->ktr_comm) ==
    (sizeof((struct thread *)NULL)->td_name));

static struct ktr_request *
ktr_getrequest_entered(struct thread *td, int type)
{
	struct ktr_request *req;
	struct proc *p = td->td_proc;
	int pm;

	mtx_lock(&ktrace_mtx);
	if (!KTRCHECK(td, type)) {
		mtx_unlock(&ktrace_mtx);
		return (NULL);
	}
	req = STAILQ_FIRST(&ktr_free);
	if (req != NULL) {
		STAILQ_REMOVE_HEAD(&ktr_free, ktr_list);
		req->ktr_header.ktr_type = type;
		if (p->p_traceflag & KTRFAC_DROP) {
			req->ktr_header.ktr_type |= KTR_DROP;
			p->p_traceflag &= ~KTRFAC_DROP;
		}
		mtx_unlock(&ktrace_mtx);
		microtime(&req->ktr_header.ktr_time);
		req->ktr_header.ktr_pid = p->p_pid;
		req->ktr_header.ktr_tid = td->td_tid;
		bcopy(td->td_name, req->ktr_header.ktr_comm,
		    sizeof(req->ktr_header.ktr_comm));
		req->ktr_buffer = NULL;
		req->ktr_header.ktr_len = 0;
	} else {
		p->p_traceflag |= KTRFAC_DROP;
		pm = print_message;
		print_message = 0;
		mtx_unlock(&ktrace_mtx);
		if (pm)
			printf("Out of ktrace request objects.\n");
	}
	return (req);
}

static struct ktr_request *
ktr_getrequest(int type)
{
	struct thread *td = curthread;
	struct ktr_request *req;

	ktrace_enter(td);
	req = ktr_getrequest_entered(td, type);
	if (req == NULL)
		ktrace_exit(td);

	return (req);
}

/*
 * Some trace generation environments don't permit direct access to VFS,
 * such as during a context switch where sleeping is not allowed.  Under these
 * circumstances, queue a request to the thread to be written asynchronously
 * later.
 */
static void
ktr_enqueuerequest(struct thread *td, struct ktr_request *req)
{

	mtx_lock(&ktrace_mtx);
	STAILQ_INSERT_TAIL(&td->td_proc->p_ktr, req, ktr_list);
	mtx_unlock(&ktrace_mtx);
}

/*
 * Drain any pending ktrace records from the per-thread queue to disk.  This
 * is used both internally before committing other records, and also on
 * system call return.  We drain all the ones we can find at the time when
 * drain is requested, but don't keep draining after that as those events
 * may be approximately "after" the current event.
 */
static void
ktr_drain(struct thread *td)
{
	struct ktr_request *queued_req;
	STAILQ_HEAD(, ktr_request) local_queue;

	ktrace_assert(td);
	sx_assert(&ktrace_sx, SX_XLOCKED);

	STAILQ_INIT(&local_queue);

	if (!STAILQ_EMPTY(&td->td_proc->p_ktr)) {
		mtx_lock(&ktrace_mtx);
		STAILQ_CONCAT(&local_queue, &td->td_proc->p_ktr);
		mtx_unlock(&ktrace_mtx);

		while ((queued_req = STAILQ_FIRST(&local_queue))) {
			STAILQ_REMOVE_HEAD(&local_queue, ktr_list);
			ktr_writerequest(td, queued_req);
			ktr_freerequest(queued_req);
		}
	}
}

/*
 * Submit a trace record for immediate commit to disk -- to be used only
 * where entering VFS is OK.  First drain any pending records that may have
 * been cached in the thread.
 */
static void
ktr_submitrequest(struct thread *td, struct ktr_request *req)
{

	ktrace_assert(td);

	sx_xlock(&ktrace_sx);
	ktr_drain(td);
	ktr_writerequest(td, req);
	ktr_freerequest(req);
	sx_xunlock(&ktrace_sx);
	ktrace_exit(td);
}

static void
ktr_freerequest(struct ktr_request *req)
{

	mtx_lock(&ktrace_mtx);
	ktr_freerequest_locked(req);
	mtx_unlock(&ktrace_mtx);
}

static void
ktr_freerequest_locked(struct ktr_request *req)
{

	mtx_assert(&ktrace_mtx, MA_OWNED);
	if (req->ktr_buffer != NULL)
		free(req->ktr_buffer, M_KTRACE);
	STAILQ_INSERT_HEAD(&ktr_free, req, ktr_list);
}

/*
 * Disable tracing for a process and release all associated resources.
 * The caller is responsible for releasing a reference on the returned
 * vnode and credentials.
 */
static void
ktr_freeproc(struct proc *p, struct ucred **uc, struct vnode **vp)
{
	struct ktr_request *req;

	PROC_LOCK_ASSERT(p, MA_OWNED);
	mtx_assert(&ktrace_mtx, MA_OWNED);
	*uc = p->p_tracecred;
	p->p_tracecred = NULL;
	if (vp != NULL)
		*vp = p->p_tracevp;
	p->p_tracevp = NULL;
	p->p_traceflag = 0;
	while ((req = STAILQ_FIRST(&p->p_ktr)) != NULL) {
		STAILQ_REMOVE_HEAD(&p->p_ktr, ktr_list);
		ktr_freerequest_locked(req);
	}
}

void
ktrsyscall(code, narg, args)
	int code, narg;
	register_t args[];
{
	struct ktr_request *req;
	struct ktr_syscall *ktp;
	size_t buflen;
	char *buf = NULL;

	buflen = sizeof(register_t) * narg;
	if (buflen > 0) {
		buf = malloc(buflen, M_KTRACE, M_WAITOK);
		bcopy(args, buf, buflen);
	}
	req = ktr_getrequest(KTR_SYSCALL);
	if (req == NULL) {
		if (buf != NULL)
			free(buf, M_KTRACE);
		return;
	}
	ktp = &req->ktr_data.ktr_syscall;
	ktp->ktr_code = code;
	ktp->ktr_narg = narg;
	if (buflen > 0) {
		req->ktr_header.ktr_len = buflen;
		req->ktr_buffer = buf;
	}
	ktr_submitrequest(curthread, req);
}

void
ktrsysret(code, error, retval)
	int code, error;
	register_t retval;
{
	struct ktr_request *req;
	struct ktr_sysret *ktp;

	req = ktr_getrequest(KTR_SYSRET);
	if (req == NULL)
		return;
	ktp = &req->ktr_data.ktr_sysret;
	ktp->ktr_code = code;
	ktp->ktr_error = error;
	ktp->ktr_retval = ((error == 0) ? retval: 0);		/* what about val2 ? */
	ktr_submitrequest(curthread, req);
}

/*
 * When a setuid process execs, disable tracing.
 *
 * XXX: We toss any pending asynchronous records.
 */
void
ktrprocexec(struct proc *p, struct ucred **uc, struct vnode **vp)
{

	PROC_LOCK_ASSERT(p, MA_OWNED);
	mtx_lock(&ktrace_mtx);
	ktr_freeproc(p, uc, vp);
	mtx_unlock(&ktrace_mtx);
}

/*
 * When a process exits, drain per-process asynchronous trace records
 * and disable tracing.
 */
void
ktrprocexit(struct thread *td)
{
	struct ktr_request *req;
	struct proc *p;
	struct ucred *cred;
	struct vnode *vp;

	p = td->td_proc;
	if (p->p_traceflag == 0)
		return;

	ktrace_enter(td);
	req = ktr_getrequest_entered(td, KTR_PROCDTOR);
	if (req != NULL)
		ktr_enqueuerequest(td, req);
	sx_xlock(&ktrace_sx);
	ktr_drain(td);
	sx_xunlock(&ktrace_sx);
	PROC_LOCK(p);
	mtx_lock(&ktrace_mtx);
	ktr_freeproc(p, &cred, &vp);
	mtx_unlock(&ktrace_mtx);
	PROC_UNLOCK(p);
	if (vp != NULL)
		vrele(vp);
	if (cred != NULL)
		crfree(cred);
	ktrace_exit(td);
}

static void
ktrprocctor_entered(struct thread *td, struct proc *p)
{
	struct ktr_proc_ctor *ktp;
	struct ktr_request *req;
	struct thread *td2;

	ktrace_assert(td);
	td2 = FIRST_THREAD_IN_PROC(p);
	req = ktr_getrequest_entered(td2, KTR_PROCCTOR);
	if (req == NULL)
		return;
	ktp = &req->ktr_data.ktr_proc_ctor;
	ktp->sv_flags = p->p_sysent->sv_flags;
	ktr_enqueuerequest(td2, req);
}

void
ktrprocctor(struct proc *p)
{
	struct thread *td = curthread;

	if ((p->p_traceflag & KTRFAC_MASK) == 0)
		return;

	ktrace_enter(td);
	ktrprocctor_entered(td, p);
	ktrace_exit(td);
}

/*
 * When a process forks, enable tracing in the new process if needed.
 */
void
ktrprocfork(struct proc *p1, struct proc *p2)
{

	PROC_LOCK(p1);
	mtx_lock(&ktrace_mtx);
	KASSERT(p2->p_tracevp == NULL, ("new process has a ktrace vnode"));
	if (p1->p_traceflag & KTRFAC_INHERIT) {
		p2->p_traceflag = p1->p_traceflag;
		if ((p2->p_tracevp = p1->p_tracevp) != NULL) {
			VREF(p2->p_tracevp);
			KASSERT(p1->p_tracecred != NULL,
			    ("ktrace vnode with no cred"));
			p2->p_tracecred = crhold(p1->p_tracecred);
		}
	}
	mtx_unlock(&ktrace_mtx);
	PROC_UNLOCK(p1);

	ktrprocctor(p2);
}

/*
 * When a thread returns, drain any asynchronous records generated by the
 * system call.
 */
void
ktruserret(struct thread *td)
{

	ktrace_enter(td);
	sx_xlock(&ktrace_sx);
	ktr_drain(td);
	sx_xunlock(&ktrace_sx);
	ktrace_exit(td);
}

void
ktrnamei(path)
	char *path;
{
	struct ktr_request *req;
	int namelen;
	char *buf = NULL;

	namelen = strlen(path);
	if (namelen > 0) {
		buf = malloc(namelen, M_KTRACE, M_WAITOK);
		bcopy(path, buf, namelen);
	}
	req = ktr_getrequest(KTR_NAMEI);
	if (req == NULL) {
		if (buf != NULL)
			free(buf, M_KTRACE);
		return;
	}
	if (namelen > 0) {
		req->ktr_header.ktr_len = namelen;
		req->ktr_buffer = buf;
	}
	ktr_submitrequest(curthread, req);
}

void
ktrsysctl(name, namelen)
	int *name;
	u_int namelen;
{
	struct ktr_request *req;
	u_int mib[CTL_MAXNAME + 2];
	char *mibname;
	size_t mibnamelen;
	int error;

	/* Lookup name of mib. */    
	KASSERT(namelen <= CTL_MAXNAME, ("sysctl MIB too long"));
	mib[0] = 0;
	mib[1] = 1;
	bcopy(name, mib + 2, namelen * sizeof(*name));
	mibnamelen = 128;
	mibname = malloc(mibnamelen, M_KTRACE, M_WAITOK);
	error = kernel_sysctl(curthread, mib, namelen + 2, mibname, &mibnamelen,
	    NULL, 0, &mibnamelen, 0);
	if (error) {
		free(mibname, M_KTRACE);
		return;
	}
	req = ktr_getrequest(KTR_SYSCTL);
	if (req == NULL) {
		free(mibname, M_KTRACE);
		return;
	}
	req->ktr_header.ktr_len = mibnamelen;
	req->ktr_buffer = mibname;
	ktr_submitrequest(curthread, req);
}

void
ktrgenio(fd, rw, uio, error)
	int fd;
	enum uio_rw rw;
	struct uio *uio;
	int error;
{
	struct ktr_request *req;
	struct ktr_genio *ktg;
	int datalen;
	char *buf;

	if (error) {
		free(uio, M_IOV);
		return;
	}
	uio->uio_offset = 0;
	uio->uio_rw = UIO_WRITE;
	datalen = MIN(uio->uio_resid, ktr_geniosize);
	buf = malloc(datalen, M_KTRACE, M_WAITOK);
	error = uiomove(buf, datalen, uio);
	free(uio, M_IOV);
	if (error) {
		free(buf, M_KTRACE);
		return;
	}
	req = ktr_getrequest(KTR_GENIO);
	if (req == NULL) {
		free(buf, M_KTRACE);
		return;
	}
	ktg = &req->ktr_data.ktr_genio;
	ktg->ktr_fd = fd;
	ktg->ktr_rw = rw;
	req->ktr_header.ktr_len = datalen;
	req->ktr_buffer = buf;
	ktr_submitrequest(curthread, req);
}

void
ktrpsig(sig, action, mask, code)
	int sig;
	sig_t action;
	sigset_t *mask;
	int code;
{
	struct thread *td = curthread;
	struct ktr_request *req;
	struct ktr_psig	*kp;

	req = ktr_getrequest(KTR_PSIG);
	if (req == NULL)
		return;
	kp = &req->ktr_data.ktr_psig;
	kp->signo = (char)sig;
	kp->action = action;
	kp->mask = *mask;
	kp->code = code;
	ktr_enqueuerequest(td, req);
	ktrace_exit(td);
}

void
ktrcsw(out, user, wmesg)
	int out, user;
	const char *wmesg;
{
	struct thread *td = curthread;
	struct ktr_request *req;
	struct ktr_csw *kc;

	req = ktr_getrequest(KTR_CSW);
	if (req == NULL)
		return;
	kc = &req->ktr_data.ktr_csw;
	kc->out = out;
	kc->user = user;
	if (wmesg != NULL)
		strlcpy(kc->wmesg, wmesg, sizeof(kc->wmesg));
	else
		bzero(kc->wmesg, sizeof(kc->wmesg));
	ktr_enqueuerequest(td, req);
	ktrace_exit(td);
}

void
ktrstruct(name, data, datalen)
	const char *name;
	void *data;
	size_t datalen;
{
	struct ktr_request *req;
	char *buf = NULL;
	size_t buflen;

	if (!data)
		datalen = 0;
	buflen = strlen(name) + 1 + datalen;
	buf = malloc(buflen, M_KTRACE, M_WAITOK);
	strcpy(buf, name);
	bcopy(data, buf + strlen(name) + 1, datalen);
	if ((req = ktr_getrequest(KTR_STRUCT)) == NULL) {
		free(buf, M_KTRACE);
		return;
	}
	req->ktr_buffer = buf;
	req->ktr_header.ktr_len = buflen;
	ktr_submitrequest(curthread, req);
}

void
ktrcapfail(type, needed, held)
	enum ktr_cap_fail_type type;
	const cap_rights_t *needed;
	const cap_rights_t *held;
{
	struct thread *td = curthread;
	struct ktr_request *req;
	struct ktr_cap_fail *kcf;

	req = ktr_getrequest(KTR_CAPFAIL);
	if (req == NULL)
		return;
	kcf = &req->ktr_data.ktr_cap_fail;
	kcf->cap_type = type;
	if (needed != NULL)
		kcf->cap_needed = *needed;
	else
		cap_rights_init(&kcf->cap_needed);
	if (held != NULL)
		kcf->cap_held = *held;
	else
		cap_rights_init(&kcf->cap_held);
	ktr_enqueuerequest(td, req);
	ktrace_exit(td);
}

void
ktrfault(vaddr, type)
	vm_offset_t vaddr;
	int type;
{
	struct thread *td = curthread;
	struct ktr_request *req;
	struct ktr_fault *kf;

	req = ktr_getrequest(KTR_FAULT);
	if (req == NULL)
		return;
	kf = &req->ktr_data.ktr_fault;
	kf->vaddr = vaddr;
	kf->type = type;
	ktr_enqueuerequest(td, req);
	ktrace_exit(td);
}

void
ktrfaultend(result)
	int result;
{
	struct thread *td = curthread;
	struct ktr_request *req;
	struct ktr_faultend *kf;

	req = ktr_getrequest(KTR_FAULTEND);
	if (req == NULL)
		return;
	kf = &req->ktr_data.ktr_faultend;
	kf->result = result;
	ktr_enqueuerequest(td, req);
	ktrace_exit(td);
}
#endif /* KTRACE */

/* Interface and common routines */

#ifndef _SYS_SYSPROTO_H_
struct ktrace_args {
	char	*fname;
	int	ops;
	int	facs;
	int	pid;
};
#endif
/* ARGSUSED */
int
sys_ktrace(td, uap)
	struct thread *td;
	register struct ktrace_args *uap;
{
#ifdef KTRACE
	register struct vnode *vp = NULL;
	register struct proc *p;
	struct pgrp *pg;
	int facs = uap->facs & ~KTRFAC_ROOT;
	int ops = KTROP(uap->ops);
	int descend = uap->ops & KTRFLAG_DESCEND;
	int nfound, ret = 0;
	int flags, error = 0;
	struct nameidata nd;
	struct ucred *cred;

	/*
	 * Need something to (un)trace.
	 */
	if (ops != KTROP_CLEARFILE && facs == 0)
		return (EINVAL);

	ktrace_enter(td);
	if (ops != KTROP_CLEAR) {
		/*
		 * an operation which requires a file argument.
		 */
		NDINIT(&nd, LOOKUP, NOFOLLOW, UIO_USERSPACE, uap->fname, td);
		flags = FREAD | FWRITE | O_NOFOLLOW;
		error = vn_open(&nd, &flags, 0, NULL);
		if (error) {
			ktrace_exit(td);
			return (error);
		}
		NDFREE(&nd, NDF_ONLY_PNBUF);
		vp = nd.ni_vp;
		VOP_UNLOCK(vp, 0);
		if (vp->v_type != VREG) {
			(void) vn_close(vp, FREAD|FWRITE, td->td_ucred, td);
			ktrace_exit(td);
			return (EACCES);
		}
	}
	/*
	 * Clear all uses of the tracefile.
	 */
	if (ops == KTROP_CLEARFILE) {
		int vrele_count;

		vrele_count = 0;
		sx_slock(&allproc_lock);
		FOREACH_PROC_IN_SYSTEM(p) {
			PROC_LOCK(p);
			if (p->p_tracevp == vp) {
				if (ktrcanset(td, p)) {
					mtx_lock(&ktrace_mtx);
					ktr_freeproc(p, &cred, NULL);
					mtx_unlock(&ktrace_mtx);
					vrele_count++;
					crfree(cred);
				} else
					error = EPERM;
			}
			PROC_UNLOCK(p);
		}
		sx_sunlock(&allproc_lock);
		if (vrele_count > 0) {
			while (vrele_count-- > 0)
				vrele(vp);
		}
		goto done;
	}
	/*
	 * do it
	 */
	sx_slock(&proctree_lock);
	if (uap->pid < 0) {
		/*
		 * by process group
		 */
		pg = pgfind(-uap->pid);
		if (pg == NULL) {
			sx_sunlock(&proctree_lock);
			error = ESRCH;
			goto done;
		}
		/*
		 * ktrops() may call vrele(). Lock pg_members
		 * by the proctree_lock rather than pg_mtx.
		 */
		PGRP_UNLOCK(pg);
		nfound = 0;
		LIST_FOREACH(p, &pg->pg_members, p_pglist) {
			PROC_LOCK(p);
			if (p->p_state == PRS_NEW ||
			    p_cansee(td, p) != 0) {
				PROC_UNLOCK(p); 
				continue;
			}
			nfound++;
			if (descend)
				ret |= ktrsetchildren(td, p, ops, facs, vp);
			else
				ret |= ktrops(td, p, ops, facs, vp);
		}
		if (nfound == 0) {
			sx_sunlock(&proctree_lock);
			error = ESRCH;
			goto done;
		}
	} else {
		/*
		 * by pid
		 */
		p = pfind(uap->pid);
		if (p == NULL)
			error = ESRCH;
		else
			error = p_cansee(td, p);
		if (error) {
			if (p != NULL)
				PROC_UNLOCK(p);
			sx_sunlock(&proctree_lock);
			goto done;
		}
		if (descend)
			ret |= ktrsetchildren(td, p, ops, facs, vp);
		else
			ret |= ktrops(td, p, ops, facs, vp);
	}
	sx_sunlock(&proctree_lock);
	if (!ret)
		error = EPERM;
done:
	if (vp != NULL)
		(void) vn_close(vp, FWRITE, td->td_ucred, td);
	ktrace_exit(td);
	return (error);
#else /* !KTRACE */
	return (ENOSYS);
#endif /* KTRACE */
}

/* ARGSUSED */
int
sys_utrace(td, uap)
	struct thread *td;
	register struct utrace_args *uap;
{

#ifdef KTRACE
	struct ktr_request *req;
	void *cp;
	int error;

	if (!KTRPOINT(td, KTR_USER))
		return (0);
	if (uap->len > KTR_USER_MAXLEN)
		return (EINVAL);
	cp = malloc(uap->len, M_KTRACE, M_WAITOK);
	error = copyin(uap->addr, cp, uap->len);
	if (error) {
		free(cp, M_KTRACE);
		return (error);
	}
	req = ktr_getrequest(KTR_USER);
	if (req == NULL) {
		free(cp, M_KTRACE);
		return (ENOMEM);
	}
	req->ktr_buffer = cp;
	req->ktr_header.ktr_len = uap->len;
	ktr_submitrequest(td, req);
	return (0);
#else /* !KTRACE */
	return (ENOSYS);
#endif /* KTRACE */
}

#ifdef KTRACE
static int
ktrops(td, p, ops, facs, vp)
	struct thread *td;
	struct proc *p;
	int ops, facs;
	struct vnode *vp;
{
	struct vnode *tracevp = NULL;
	struct ucred *tracecred = NULL;

	PROC_LOCK_ASSERT(p, MA_OWNED);
	if (!ktrcanset(td, p)) {
		PROC_UNLOCK(p);
		return (0);
	}
	if (p->p_flag & P_WEXIT) {
		/* If the process is exiting, just ignore it. */
		PROC_UNLOCK(p);
		return (1);
	}
	mtx_lock(&ktrace_mtx);
	if (ops == KTROP_SET) {
		if (p->p_tracevp != vp) {
			/*
			 * if trace file already in use, relinquish below
			 */
			tracevp = p->p_tracevp;
			VREF(vp);
			p->p_tracevp = vp;
		}
		if (p->p_tracecred != td->td_ucred) {
			tracecred = p->p_tracecred;
			p->p_tracecred = crhold(td->td_ucred);
		}
		p->p_traceflag |= facs;
		if (priv_check(td, PRIV_KTRACE) == 0)
			p->p_traceflag |= KTRFAC_ROOT;
	} else {
		/* KTROP_CLEAR */
		if (((p->p_traceflag &= ~facs) & KTRFAC_MASK) == 0)
			/* no more tracing */
			ktr_freeproc(p, &tracecred, &tracevp);
	}
	mtx_unlock(&ktrace_mtx);
	if ((p->p_traceflag & KTRFAC_MASK) != 0)
		ktrprocctor_entered(td, p);
	PROC_UNLOCK(p);
	if (tracevp != NULL)
		vrele(tracevp);
	if (tracecred != NULL)
		crfree(tracecred);

	return (1);
}

static int
ktrsetchildren(td, top, ops, facs, vp)
	struct thread *td;
	struct proc *top;
	int ops, facs;
	struct vnode *vp;
{
	register struct proc *p;
	register int ret = 0;

	p = top;
	PROC_LOCK_ASSERT(p, MA_OWNED);
	sx_assert(&proctree_lock, SX_LOCKED);
	for (;;) {
		ret |= ktrops(td, p, ops, facs, vp);
		/*
		 * If this process has children, descend to them next,
		 * otherwise do any siblings, and if done with this level,
		 * follow back up the tree (but not past top).
		 */
		if (!LIST_EMPTY(&p->p_children))
			p = LIST_FIRST(&p->p_children);
		else for (;;) {
			if (p == top)
				return (ret);
			if (LIST_NEXT(p, p_sibling)) {
				p = LIST_NEXT(p, p_sibling);
				break;
			}
			p = p->p_pptr;
		}
		PROC_LOCK(p);
	}
	/*NOTREACHED*/
}

static void
ktr_writerequest(struct thread *td, struct ktr_request *req)
{
	struct ktr_header *kth;
	struct vnode *vp;
	struct proc *p;
	struct ucred *cred;
	struct uio auio;
	struct iovec aiov[3];
	struct mount *mp;
	int datalen, buflen, vrele_count;
	int error;

	/*
	 * We hold the vnode and credential for use in I/O in case ktrace is
	 * disabled on the process as we write out the request.
	 *
	 * XXXRW: This is not ideal: we could end up performing a write after
	 * the vnode has been closed.
	 */
	mtx_lock(&ktrace_mtx);
	vp = td->td_proc->p_tracevp;
	cred = td->td_proc->p_tracecred;

	/*
	 * If vp is NULL, the vp has been cleared out from under this
	 * request, so just drop it.  Make sure the credential and vnode are
	 * in sync: we should have both or neither.
	 */
	if (vp == NULL) {
		KASSERT(cred == NULL, ("ktr_writerequest: cred != NULL"));
		mtx_unlock(&ktrace_mtx);
		return;
	}
	VREF(vp);
	KASSERT(cred != NULL, ("ktr_writerequest: cred == NULL"));
	crhold(cred);
	mtx_unlock(&ktrace_mtx);

	kth = &req->ktr_header;
	KASSERT(((u_short)kth->ktr_type & ~KTR_DROP) <
	    sizeof(data_lengths) / sizeof(data_lengths[0]),
	    ("data_lengths array overflow"));
	datalen = data_lengths[(u_short)kth->ktr_type & ~KTR_DROP];
	buflen = kth->ktr_len;
	auio.uio_iov = &aiov[0];
	auio.uio_offset = 0;
	auio.uio_segflg = UIO_SYSSPACE;
	auio.uio_rw = UIO_WRITE;
	aiov[0].iov_base = (caddr_t)kth;
	aiov[0].iov_len = sizeof(struct ktr_header);
	auio.uio_resid = sizeof(struct ktr_header);
	auio.uio_iovcnt = 1;
	auio.uio_td = td;
	if (datalen != 0) {
		aiov[1].iov_base = (caddr_t)&req->ktr_data;
		aiov[1].iov_len = datalen;
		auio.uio_resid += datalen;
		auio.uio_iovcnt++;
		kth->ktr_len += datalen;
	}
	if (buflen != 0) {
		KASSERT(req->ktr_buffer != NULL, ("ktrace: nothing to write"));
		aiov[auio.uio_iovcnt].iov_base = req->ktr_buffer;
		aiov[auio.uio_iovcnt].iov_len = buflen;
		auio.uio_resid += buflen;
		auio.uio_iovcnt++;
	}

	vn_start_write(vp, &mp, V_WAIT);
	vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
#ifdef MAC
	error = mac_vnode_check_write(cred, NOCRED, vp);
	if (error == 0)
#endif
		error = VOP_WRITE(vp, &auio, IO_UNIT | IO_APPEND, cred);
	VOP_UNLOCK(vp, 0);
	vn_finished_write(mp);
	crfree(cred);
	if (!error) {
		vrele(vp);
		return;
	}

	/*
	 * If error encountered, give up tracing on this vnode.  We defer
	 * all the vrele()'s on the vnode until after we are finished walking
	 * the various lists to avoid needlessly holding locks.
	 * NB: at this point we still hold the vnode reference that must
	 * not go away as we need the valid vnode to compare with. Thus let
	 * vrele_count start at 1 and the reference will be freed
	 * by the loop at the end after our last use of vp.
	 */
	log(LOG_NOTICE, "ktrace write failed, errno %d, tracing stopped\n",
	    error);
	vrele_count = 1;
	/*
	 * First, clear this vnode from being used by any processes in the
	 * system.
	 * XXX - If one process gets an EPERM writing to the vnode, should
	 * we really do this?  Other processes might have suitable
	 * credentials for the operation.
	 */
	cred = NULL;
	sx_slock(&allproc_lock);
	FOREACH_PROC_IN_SYSTEM(p) {
		PROC_LOCK(p);
		if (p->p_tracevp == vp) {
			mtx_lock(&ktrace_mtx);
			ktr_freeproc(p, &cred, NULL);
			mtx_unlock(&ktrace_mtx);
			vrele_count++;
		}
		PROC_UNLOCK(p);
		if (cred != NULL) {
			crfree(cred);
			cred = NULL;
		}
	}
	sx_sunlock(&allproc_lock);

	while (vrele_count-- > 0)
		vrele(vp);
}

/*
 * Return true if caller has permission to set the ktracing state
 * of target.  Essentially, the target can't possess any
 * more permissions than the caller.  KTRFAC_ROOT signifies that
 * root previously set the tracing status on the target process, and
 * so, only root may further change it.
 */
static int
ktrcanset(td, targetp)
	struct thread *td;
	struct proc *targetp;
{

	PROC_LOCK_ASSERT(targetp, MA_OWNED);
	if (targetp->p_traceflag & KTRFAC_ROOT &&
	    priv_check(td, PRIV_KTRACE))
		return (0);

	if (p_candebug(td, targetp) != 0)
		return (0);

	return (1);
}

#endif /* KTRACE */
OpenPOWER on IntegriCloud