summaryrefslogtreecommitdiff
path: root/sys/kern/kern_time.c
blob: 2e83b07ffb1444729189d42cd5741f9a0a3c21d9 (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
/*	$OpenBSD: kern_time.c,v 1.75 2012/05/24 07:17:42 guenther Exp $	*/
/*	$NetBSD: kern_time.c,v 1.20 1996/02/18 11:57:06 fvdl Exp $	*/

/*
 * Copyright (c) 1982, 1986, 1989, 1993
 *	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. 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_time.c	8.4 (Berkeley) 5/26/95
 */

#include <sys/param.h>
#include <sys/resourcevar.h>
#include <sys/kernel.h>
#include <sys/systm.h>
#include <sys/proc.h>
#include <sys/ktrace.h>
#include <sys/vnode.h>
#include <sys/signalvar.h>
#ifdef __HAVE_TIMECOUNTER
#include <sys/timetc.h>
#endif

#include <sys/mount.h>
#include <sys/syscallargs.h>

#include <machine/cpu.h>

#ifdef __HAVE_TIMECOUNTER
struct timeval adjtimedelta;		/* unapplied time correction */
#else
int	tickdelta;			/* current clock skew, us. per tick */
long	timedelta;			/* unapplied time correction, us. */
long	bigadj = 1000000;		/* use 10x skew above bigadj us. */
int64_t	ntp_tick_permanent;
int64_t	ntp_tick_acc;
#endif

void	itimerround(struct timeval *);

/* 
 * Time of day and interval timer support.
 *
 * These routines provide the kernel entry points to get and set
 * the time-of-day and per-process interval timers.  Subroutines
 * here provide support for adding and subtracting timeval structures
 * and decrementing interval timers, optionally reloading the interval
 * timers when they expire.
 */

/* This function is used by clock_settime and settimeofday */
#ifdef __HAVE_TIMECOUNTER
int
settime(struct timespec *ts)
{
	struct timespec now;

	/*
	 * Adjtime in progress is meaningless or harmful after
	 * setting the clock. Cancel adjtime and then set new time.
	 */
	adjtimedelta.tv_usec = 0;
	adjtimedelta.tv_sec = 0;

	/*
	 * Don't allow the time to be set forward so far it will wrap
	 * and become negative, thus allowing an attacker to bypass
	 * the next check below.  The cutoff is 1 year before rollover
	 * occurs, so even if the attacker uses adjtime(2) to move
	 * the time past the cutoff, it will take a very long time
	 * to get to the wrap point.
	 *
	 * XXX: we check against INT_MAX since on 64-bit
	 *	platforms, sizeof(int) != sizeof(long) and
	 *	time_t is 32 bits even when atv.tv_sec is 64 bits.
	 */
	if (ts->tv_sec > INT_MAX - 365*24*60*60) {
		printf("denied attempt to set clock forward to %ld\n",
		    ts->tv_sec);
		return (EPERM);
	}
	/*
	 * If the system is secure, we do not allow the time to be
	 * set to an earlier value (it may be slowed using adjtime,
	 * but not set back). This feature prevent interlopers from
	 * setting arbitrary time stamps on files.
	 */
	nanotime(&now);
	if (securelevel > 1 && timespeccmp(ts, &now, <)) {
		printf("denied attempt to set clock back %ld seconds\n",
		    now.tv_sec - ts->tv_sec);
		return (EPERM);
	}

	tc_setrealtimeclock(ts);
	resettodr();

	return (0);
}
#else
int
settime(struct timespec *ts)
{
	struct timeval delta, tvv, *tv;
	int s;

	/* XXX - Ugh. */
	tv = &tvv;
	tvv.tv_sec = ts->tv_sec;
	tvv.tv_usec = ts->tv_nsec / 1000;

	/*
	 * Don't allow the time to be set forward so far it will wrap
	 * and become negative, thus allowing an attacker to bypass
	 * the next check below.  The cutoff is 1 year before rollover
	 * occurs, so even if the attacker uses adjtime(2) to move
	 * the time past the cutoff, it will take a very long time
	 * to get to the wrap point.
	 *
	 * XXX: we check against INT_MAX since on 64-bit
	 *	platforms, sizeof(int) != sizeof(long) and
	 *	time_t is 32 bits even when atv.tv_sec is 64 bits.
	 */
	if (tv->tv_sec > INT_MAX - 365*24*60*60) {
		printf("denied attempt to set clock forward to %ld\n",
		    tv->tv_sec);
		return (EPERM);
	}
	/*
	 * If the system is secure, we do not allow the time to be
	 * set to an earlier value (it may be slowed using adjtime,
	 * but not set back). This feature prevent interlopers from
	 * setting arbitrary time stamps on files.
	 */
	if (securelevel > 1 && timercmp(tv, &time, <)) {
		printf("denied attempt to set clock back %ld seconds\n",
		    time_second - tv->tv_sec);
		return (EPERM);
	}

	/* WHAT DO WE DO ABOUT PENDING REAL-TIME TIMEOUTS??? */
	s = splclock();
	timersub(tv, &time, &delta);
	time = *tv;
	timeradd(&boottime, &delta, &boottime);

	/*
	 * Adjtime in progress is meaningless or harmful after
	 * setting the clock.
	 */
	tickdelta = 0;
	timedelta = 0;

	splx(s);
	resettodr();

	return (0);
}
#endif

int
clock_gettime(struct proc *p, clockid_t clock_id, struct timespec *tp)
{
	struct timeval tv;

	switch (clock_id) {
	case CLOCK_REALTIME:
		nanotime(tp);
		break;
	case CLOCK_MONOTONIC:
		nanouptime(tp);
		break;
	case CLOCK_PROF:
		microuptime(&tv);
		timersub(&tv, &curcpu()->ci_schedstate.spc_runtime, &tv);
		timeradd(&tv, &p->p_rtime, &tv);
		tp->tv_sec = tv.tv_sec;
		tp->tv_nsec = tv.tv_usec * 1000;
		break;
	default:
		return (EINVAL);
	}
	return (0);
}

/* ARGSUSED */
int
sys_clock_gettime(struct proc *p, void *v, register_t *retval)
{
	struct sys_clock_gettime_args /* {
		syscallarg(clockid_t) clock_id;
		syscallarg(struct timespec *) tp;
	} */ *uap = v;
	struct timespec ats;
	int error;

	if ((error = clock_gettime(p, SCARG(uap, clock_id), &ats)) != 0)
		return (error);

	error = copyout(&ats, SCARG(uap, tp), sizeof(ats));
#ifdef KTRACE
	if (error == 0 && KTRPOINT(p, KTR_STRUCT)) {
		KERNEL_LOCK();
		ktrabstimespec(p, &ats);
		KERNEL_UNLOCK();
	}
#endif
	return (error);
}

/* ARGSUSED */
int
sys_clock_settime(struct proc *p, void *v, register_t *retval)
{
	struct sys_clock_settime_args /* {
		syscallarg(clockid_t) clock_id;
		syscallarg(const struct timespec *) tp;
	} */ *uap = v;
	struct timespec ats;
	clockid_t clock_id;
	int error;

	if ((error = suser(p, 0)) != 0)
		return (error);

	if ((error = copyin(SCARG(uap, tp), &ats, sizeof(ats))) != 0)
		return (error);

	clock_id = SCARG(uap, clock_id);
	switch (clock_id) {
	case CLOCK_REALTIME:
		if ((error = settime(&ats)) != 0)
			return (error);
		break;
	default:	/* Other clocks are read-only */
		return (EINVAL);
	}

	return (0);
}

int
sys_clock_getres(struct proc *p, void *v, register_t *retval)
{
	struct sys_clock_getres_args /* {
		syscallarg(clockid_t) clock_id;
		syscallarg(struct timespec *) tp;
	} */ *uap = v;
	clockid_t clock_id;
	struct timespec ts;
	int error = 0;

	clock_id = SCARG(uap, clock_id);
	switch (clock_id) {
	case CLOCK_REALTIME:
	case CLOCK_MONOTONIC:
		ts.tv_sec = 0;
		ts.tv_nsec = 1000000000 / hz;
		break;
	default:
		return (EINVAL);
	}

	if (SCARG(uap, tp)) {
		error = copyout(&ts, SCARG(uap, tp), sizeof (ts));
#ifdef KTRACE
		if (error == 0 && KTRPOINT(p, KTR_STRUCT)) {
			KERNEL_LOCK();
			ktrreltimespec(p, &ts);
			KERNEL_UNLOCK();
		}
#endif
	}

	return error;
}

/* ARGSUSED */
int
sys_nanosleep(struct proc *p, void *v, register_t *retval)
{
	static int nanowait;
	struct sys_nanosleep_args/* {
		syscallarg(const struct timespec *) rqtp;
		syscallarg(struct timespec *) rmtp;
	} */ *uap = v;
	struct timespec rqt, rmt;
	struct timespec sts, ets;
	struct timespec *rmtp;
	struct timeval tv;
	int error, error1;

	rmtp = SCARG(uap, rmtp);
	error = copyin(SCARG(uap, rqtp), &rqt, sizeof(struct timespec));
	if (error)
		return (error);
#ifdef KTRACE
        if (KTRPOINT(p, KTR_STRUCT)) {
		KERNEL_LOCK();
		ktrreltimespec(p, &rqt);
		KERNEL_UNLOCK();
	}
#endif

	TIMESPEC_TO_TIMEVAL(&tv, &rqt);
	if (itimerfix(&tv))
		return (EINVAL);

	if (rmtp)
		getnanouptime(&sts);

	error = tsleep(&nanowait, PWAIT | PCATCH, "nanosleep",
	    MAX(1, tvtohz(&tv)));
	if (error == ERESTART)
		error = EINTR;
	if (error == EWOULDBLOCK)
		error = 0;

	if (rmtp) {
		getnanouptime(&ets);

		timespecsub(&ets, &sts, &sts);
		timespecsub(&rqt, &sts, &rmt);

		if (rmt.tv_sec < 0)
			timespecclear(&rmt);

		error1 = copyout(&rmt, rmtp, sizeof(rmt));
		if (error1 != 0)
			error = error1;
#ifdef KTRACE
		if (error1 == 0 && KTRPOINT(p, KTR_STRUCT)) {
			KERNEL_LOCK();
			ktrreltimespec(p, &rmt);
			KERNEL_UNLOCK();
		}
#endif
	}

	return error;
}

/* ARGSUSED */
int
sys_gettimeofday(struct proc *p, void *v, register_t *retval)
{
	struct sys_gettimeofday_args /* {
		syscallarg(struct timeval *) tp;
		syscallarg(struct timezone *) tzp;
	} */ *uap = v;
	struct timeval atv;
	struct timeval *tp;
	struct timezone *tzp;
	int error = 0;

	tp = SCARG(uap, tp);
	tzp = SCARG(uap, tzp);

	if (tp) {
		microtime(&atv);
		if ((error = copyout(&atv, tp, sizeof (atv))))
			return (error);
#ifdef KTRACE
		if (KTRPOINT(p, KTR_STRUCT)) {
			KERNEL_LOCK();
			ktrabstimeval(p, &atv);
			KERNEL_UNLOCK();
		}
#endif
	}
	if (tzp)
		error = copyout(&tz, tzp, sizeof (tz));
	return (error);
}

/* ARGSUSED */
int
sys_settimeofday(struct proc *p, void *v, register_t *retval)
{
	struct sys_settimeofday_args /* {
		syscallarg(const struct timeval *) tv;
		syscallarg(const struct timezone *) tzp;
	} */ *uap = v;
	struct timezone atz;
	struct timeval atv;
	const struct timeval *tv;
	const struct timezone *tzp;
	int error;

	tv = SCARG(uap, tv);
	tzp = SCARG(uap, tzp);

	if ((error = suser(p, 0)))
		return (error);
	/* Verify all parameters before changing time. */
	if (tv && (error = copyin(tv, &atv, sizeof(atv))))
		return (error);
	if (tzp && (error = copyin(tzp, &atz, sizeof(atz))))
		return (error);
	if (tv) {
		struct timespec ts;

		TIMEVAL_TO_TIMESPEC(&atv, &ts);
		if ((error = settime(&ts)) != 0)
			return (error);
	}
	if (tzp)
		tz = atz;
	return (0);
}

/* ARGSUSED */
int
sys_adjfreq(struct proc *p, void *v, register_t *retval)
{
	struct sys_adjfreq_args /* {
		syscallarg(const int64_t *) freq;
		syscallarg(int64_t *) oldfreq;
	} */ *uap = v;
	int error;
	int64_t f;
	const int64_t *freq = SCARG(uap, freq);
	int64_t *oldfreq = SCARG(uap, oldfreq);
#ifndef __HAVE_TIMECOUNTER
	int s;

	if (oldfreq) {
		f = ntp_tick_permanent * hz;
		if ((error = copyout(&f, oldfreq, sizeof(int64_t))))
			return (error);
	}
	if (freq) {
		if ((error = suser(p, 0)))
			return (error);
		if ((error = copyin(freq, &f, sizeof(int64_t))))
			return (error);
		s = splclock();
		ntp_tick_permanent = f / hz;
		splx(s);
	}
#else
	if (oldfreq) {
		if ((error = tc_adjfreq(&f, NULL)))
			return (error);
		if ((error = copyout(&f, oldfreq, sizeof(f))))
			return (error);
	}
	if (freq) {
		if ((error = suser(p, 0)))
			return (error);
		if ((error = copyin(freq, &f, sizeof(f))))
			return (error);
		if ((error = tc_adjfreq(NULL, &f)))
			return (error);
	}
#endif
	return (0);
}

/* ARGSUSED */
int
sys_adjtime(struct proc *p, void *v, register_t *retval)
{
	struct sys_adjtime_args /* {
		syscallarg(const struct timeval *) delta;
		syscallarg(struct timeval *) olddelta;
	} */ *uap = v;
	const struct timeval *delta = SCARG(uap, delta);
	struct timeval *olddelta = SCARG(uap, olddelta);
#ifdef __HAVE_TIMECOUNTER
	int error;

	if (olddelta)
		if ((error = copyout(&adjtimedelta, olddelta,
		    sizeof(struct timeval))))
			return (error);

	if (delta) {
		if ((error = suser(p, 0)))
			return (error);

		if ((error = copyin(delta, &adjtimedelta,
		    sizeof(struct timeval))))
			return (error);
	}

	/* Normalize the correction. */
	while (adjtimedelta.tv_usec >= 1000000) {
		adjtimedelta.tv_usec -= 1000000;
		adjtimedelta.tv_sec += 1;
	}
	while (adjtimedelta.tv_usec < 0) {
		adjtimedelta.tv_usec += 1000000;
		adjtimedelta.tv_sec -= 1;
	}
	return (0);
#else
	struct timeval atv;
	long ndelta, ntickdelta, odelta;
	int s, error;

	if (!delta) {
		s = splclock();
		odelta = timedelta;
		splx(s);
		goto out;
	}
	if ((error = suser(p, 0)))
		return (error);
	if ((error = copyin(delta, &atv, sizeof(struct timeval))))
		return (error);

	/*
	 * Compute the total correction and the rate at which to apply it.
	 * Round the adjustment down to a whole multiple of the per-tick
	 * delta, so that after some number of incremental changes in
	 * hardclock(), tickdelta will become zero, lest the correction
	 * overshoot and start taking us away from the desired final time.
	 */
	if (atv.tv_sec > LONG_MAX / 1000000L)
		ndelta = LONG_MAX;
	else if (atv.tv_sec < LONG_MIN / 1000000L)
		ndelta = LONG_MIN;
	else {
		ndelta = atv.tv_sec * 1000000L;
		odelta = ndelta;
		ndelta += atv.tv_usec;
		if (atv.tv_usec > 0 && ndelta <= odelta)
			ndelta = LONG_MAX;
		else if (atv.tv_usec < 0 && ndelta >= odelta)
			ndelta = LONG_MIN;
	}

	if (ndelta > bigadj || ndelta < -bigadj)
		ntickdelta = 10 * tickadj;
	else
		ntickdelta = tickadj;
	if (ndelta % ntickdelta)
		ndelta = ndelta / ntickdelta * ntickdelta;

	/*
	 * To make hardclock()'s job easier, make the per-tick delta negative
	 * if we want time to run slower; then hardclock can simply compute
	 * tick + tickdelta, and subtract tickdelta from timedelta.
	 */
	if (ndelta < 0)
		ntickdelta = -ntickdelta;
	s = splclock();
	odelta = timedelta;
	timedelta = ndelta;
	tickdelta = ntickdelta;
	splx(s);

out:
	if (olddelta) {
		atv.tv_sec = odelta / 1000000;
		atv.tv_usec = odelta % 1000000;
		if ((error = copyout(&atv, olddelta, sizeof(struct timeval))))
			return (error);
	}
	return (0);
#endif
}


/*
 * Get value of an interval timer.  The process virtual and
 * profiling virtual time timers are kept internally in the
 * way they are specified externally: in time until they expire.
 *
 * The real time interval timer's it_value, in contast, is kept as an 
 * absolute time rather than as a delta, so that it is easy to keep
 * periodic real-time signals from drifting.
 *
 * Virtual time timers are processed in the hardclock() routine of
 * kern_clock.c.  The real time timer is processed by a timeout
 * routine, called from the softclock() routine.  Since a callout
 * may be delayed in real time due to interrupt processing in the system,
 * it is possible for the real time timeout routine (realitexpire, given below),
 * to be delayed in real time past when it is supposed to occur.  It
 * does not suffice, therefore, to reload the real timer .it_value from the
 * real time timers .it_interval.  Rather, we compute the next time in
 * absolute time the timer should go off.
 */
/* ARGSUSED */
int
sys_getitimer(struct proc *p, void *v, register_t *retval)
{
	struct sys_getitimer_args /* {
		syscallarg(int) which;
		syscallarg(struct itimerval *) itv;
	} */ *uap = v;
	struct itimerval aitv;
	int s;
	int which;

	which = SCARG(uap, which);

	if (which < ITIMER_REAL || which > ITIMER_PROF)
		return (EINVAL);
	s = splclock();
	aitv = p->p_p->ps_timer[which];

	if (which == ITIMER_REAL) {
		struct timeval now;

		getmicrouptime(&now);
		/*
		 * Convert from absolute to relative time in .it_value
		 * part of real time timer.  If time for real time timer
		 * has passed return 0, else return difference between
		 * current time and time for the timer to go off.
		 */
		if (timerisset(&aitv.it_value)) {
			if (timercmp(&aitv.it_value, &now, <))
				timerclear(&aitv.it_value);
			else
				timersub(&aitv.it_value, &now,
				    &aitv.it_value);
		}
	}
	splx(s);
	return (copyout(&aitv, SCARG(uap, itv), sizeof (struct itimerval)));
}

/* ARGSUSED */
int
sys_setitimer(struct proc *p, void *v, register_t *retval)
{
	struct sys_setitimer_args /* {
		syscallarg(int) which;
		syscallarg(const struct itimerval *) itv;
		syscallarg(struct itimerval *) oitv;
	} */ *uap = v;
	struct sys_getitimer_args getargs;
	struct itimerval aitv;
	const struct itimerval *itvp;
	struct itimerval *oitv;
	struct process *pr = p->p_p;
	int error;
	int timo;
	int which;

	which = SCARG(uap, which);
	oitv = SCARG(uap, oitv);

	if (which < ITIMER_REAL || which > ITIMER_PROF)
		return (EINVAL);
	itvp = SCARG(uap, itv);
	if (itvp && (error = copyin((void *)itvp, (void *)&aitv,
	    sizeof(struct itimerval))))
		return (error);
	if (oitv != NULL) {
		SCARG(&getargs, which) = which;
		SCARG(&getargs, itv) = oitv;
		if ((error = sys_getitimer(p, &getargs, retval)))
			return (error);
	}
	if (itvp == 0)
		return (0);
	if (itimerfix(&aitv.it_value) || itimerfix(&aitv.it_interval))
		return (EINVAL);
	if (which == ITIMER_REAL) {
		struct timeval ctv;

		timeout_del(&pr->ps_realit_to);
		getmicrouptime(&ctv);
		if (timerisset(&aitv.it_value)) {
			timo = tvtohz(&aitv.it_value);
			timeout_add(&pr->ps_realit_to, timo);
			timeradd(&aitv.it_value, &ctv, &aitv.it_value);
		}
		pr->ps_timer[ITIMER_REAL] = aitv;
	} else {
		int s;

		itimerround(&aitv.it_interval);
		s = splclock();
		pr->ps_timer[which] = aitv;
		if (which == ITIMER_VIRTUAL)
			timeout_del(&pr->ps_virt_to);
		if (which == ITIMER_PROF)
			timeout_del(&pr->ps_prof_to);
		splx(s);
	}

	return (0);
}

/*
 * Real interval timer expired:
 * send process whose timer expired an alarm signal.
 * If time is not set up to reload, then just return.
 * Else compute next time timer should go off which is > current time.
 * This is where delay in processing this timeout causes multiple
 * SIGALRM calls to be compressed into one.
 */
void
realitexpire(void *arg)
{
	struct process *pr = arg;
	struct itimerval *tp = &pr->ps_timer[ITIMER_REAL];

	psignal(pr->ps_mainproc, SIGALRM);
	if (!timerisset(&tp->it_interval)) {
		timerclear(&tp->it_value);
		return;
	}
	for (;;) {
		struct timeval ctv, ntv;
		int timo;

		timeradd(&tp->it_value, &tp->it_interval, &tp->it_value);
		getmicrouptime(&ctv);
		if (timercmp(&tp->it_value, &ctv, >)) {
			ntv = tp->it_value;
			timersub(&ntv, &ctv, &ntv);
			timo = tvtohz(&ntv) - 1;
			if (timo <= 0)
				timo = 1;
			if ((pr->ps_flags & PS_EXITING) == 0)
				timeout_add(&pr->ps_realit_to, timo);
			return;
		}
	}
}

/*
 * Check that a timespec value is legit
 */
int
timespecfix(struct timespec *ts)
{
	if (ts->tv_sec < 0 || ts->tv_sec > 100000000 ||
	    ts->tv_nsec < 0 || ts->tv_nsec >= 1000000000)
		return (EINVAL);
	return (0);
}

/*
 * Check that a proposed value to load into the .it_value or
 * .it_interval part of an interval timer is acceptable.
 */
int
itimerfix(struct timeval *tv)
{

	if (tv->tv_sec < 0 || tv->tv_sec > 100000000 ||
	    tv->tv_usec < 0 || tv->tv_usec >= 1000000)
		return (EINVAL);

	if (tv->tv_sec == 0 && tv->tv_usec != 0 && tv->tv_usec < tick)
		tv->tv_usec = tick;

	return (0);
}

/*
 * Nonzero timer interval smaller than the resolution of the
 * system clock are rounded up.
 */
void
itimerround(struct timeval *tv)
{
	if (tv->tv_sec == 0 && tv->tv_usec != 0 && tv->tv_usec < tick)
		tv->tv_usec = tick;
}

/*
 * Decrement an interval timer by a specified number
 * of microseconds, which must be less than a second,
 * i.e. < 1000000.  If the timer expires, then reload
 * it.  In this case, carry over (usec - old value) to
 * reduce the value reloaded into the timer so that
 * the timer does not drift.  This routine assumes
 * that it is called in a context where the timers
 * on which it is operating cannot change in value.
 */
int
itimerdecr(struct itimerval *itp, int usec)
{

	if (itp->it_value.tv_usec < usec) {
		if (itp->it_value.tv_sec == 0) {
			/* expired, and already in next interval */
			usec -= itp->it_value.tv_usec;
			goto expire;
		}
		itp->it_value.tv_usec += 1000000;
		itp->it_value.tv_sec--;
	}
	itp->it_value.tv_usec -= usec;
	usec = 0;
	if (timerisset(&itp->it_value))
		return (1);
	/* expired, exactly at end of interval */
expire:
	if (timerisset(&itp->it_interval)) {
		itp->it_value = itp->it_interval;
		itp->it_value.tv_usec -= usec;
		if (itp->it_value.tv_usec < 0) {
			itp->it_value.tv_usec += 1000000;
			itp->it_value.tv_sec--;
		}
	} else
		itp->it_value.tv_usec = 0;		/* sec is already 0 */
	return (0);
}

/*
 * ratecheck(): simple time-based rate-limit checking.  see ratecheck(9)
 * for usage and rationale.
 */
int
ratecheck(struct timeval *lasttime, const struct timeval *mininterval)
{
	struct timeval tv, delta;
	int rv = 0;

	getmicrouptime(&tv);

	timersub(&tv, lasttime, &delta);

	/*
	 * check for 0,0 is so that the message will be seen at least once,
	 * even if interval is huge.
	 */
	if (timercmp(&delta, mininterval, >=) ||
	    (lasttime->tv_sec == 0 && lasttime->tv_usec == 0)) {
		*lasttime = tv;
		rv = 1;
	}

	return (rv);
}

/*
 * ppsratecheck(): packets (or events) per second limitation.
 */
int
ppsratecheck(struct timeval *lasttime, int *curpps, int maxpps)
{
	struct timeval tv, delta;
	int rv;

	microuptime(&tv);

	timersub(&tv, lasttime, &delta);

	/*
	 * check for 0,0 is so that the message will be seen at least once.
	 * if more than one second have passed since the last update of
	 * lasttime, reset the counter.
	 *
	 * we do increment *curpps even in *curpps < maxpps case, as some may
	 * try to use *curpps for stat purposes as well.
	 */
	if (maxpps == 0)
		rv = 0;
	else if ((lasttime->tv_sec == 0 && lasttime->tv_usec == 0) ||
	    delta.tv_sec >= 1) {
		*lasttime = tv;
		*curpps = 0;
		rv = 1;
	} else if (maxpps < 0)
		rv = 1;
	else if (*curpps < maxpps)
		rv = 1;
	else
		rv = 0;

#if 1 /*DIAGNOSTIC?*/
	/* be careful about wrap-around */
	if (*curpps + 1 > *curpps)
		*curpps = *curpps + 1;
#else
	/*
	 * assume that there's not too many calls to this function.
	 * not sure if the assumption holds, as it depends on *caller's*
	 * behavior, not the behavior of this function.
	 * IMHO it is wrong to make assumption on the caller's behavior,
	 * so the above #if is #if 1, not #ifdef DIAGNOSTIC.
	 */
	*curpps = *curpps + 1;
#endif

	return (rv);
}