ntp.c 12.6 KB
Newer Older
1 2 3 4 5 6 7
/*
 * NTP state machine interfaces and logic.
 *
 * This code was mainly moved from kernel/timer.c and kernel/time.c
 * Please see those files for relevant copyright info and historical
 * changelogs.
 */
A
Alexey Dobriyan 已提交
8
#include <linux/capability.h>
R
Roman Zippel 已提交
9
#include <linux/clocksource.h>
10
#include <linux/workqueue.h>
11 12 13 14 15 16
#include <linux/hrtimer.h>
#include <linux/jiffies.h>
#include <linux/math64.h>
#include <linux/timex.h>
#include <linux/time.h>
#include <linux/mm.h>
17

18
/*
19
 * NTP timekeeping variables:
20 21
 */

22 23 24 25 26
/* USER_HZ period (usecs): */
unsigned long			tick_usec = TICK_USEC;

/* ACTHZ period (nsecs): */
unsigned long			tick_nsec;
R
Roman Zippel 已提交
27

28 29 30 31 32
u64				tick_length;
static u64			tick_length_base;

static struct hrtimer		leap_timer;

33
#define MAX_TICKADJ		500LL		/* usecs */
34
#define MAX_TICKADJ_SCALED \
35
	(((MAX_TICKADJ * NSEC_PER_USEC) << NTP_SCALE_SHIFT) / NTP_INTERVAL_FREQ)
36 37 38 39

/*
 * phase-lock loop variables
 */
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

/*
 * clock synchronization status
 *
 * (TIME_ERROR prevents overwriting the CMOS clock)
 */
static int			time_state = TIME_OK;

/* clock status bits:							*/
int				time_status = STA_UNSYNC;

/* TAI offset (secs):							*/
static long			time_tai;

/* time adjustment (nsecs):						*/
static s64			time_offset;

/* pll time constant:							*/
static long			time_constant = 2;

/* maximum error (usecs):						*/
long				time_maxerror = NTP_PHASE_LIMIT;

/* estimated error (usecs):						*/
long				time_esterror = NTP_PHASE_LIMIT;

/* frequency offset (scaled nsecs/secs):				*/
static s64			time_freq;

/* time at last adjustment (secs):					*/
static long			time_reftime;

long				time_adjust;

74 75
/* constant (boot-param configurable) NTP tick adjustment (upscaled)	*/
static s64			ntp_tick_adj;
76 77 78 79

/*
 * NTP methods:
 */
80

81 82 83 84
/*
 * Update (tick_length, tick_length_base, tick_nsec), based
 * on (tick_usec, ntp_tick_adj, time_freq):
 */
85 86
static void ntp_update_frequency(void)
{
87
	u64 second_length;
88
	u64 new_base;
89 90 91 92

	second_length		 = (u64)(tick_usec * NSEC_PER_USEC * USER_HZ)
						<< NTP_SCALE_SHIFT;

93
	second_length		+= ntp_tick_adj;
94
	second_length		+= time_freq;
95

96
	tick_nsec		 = div_u64(second_length, HZ) >> NTP_SCALE_SHIFT;
97
	new_base		 = div_u64(second_length, NTP_INTERVAL_FREQ);
98 99 100

	/*
	 * Don't wait for the next second_overflow, apply
101
	 * the change to the tick length immediately:
102
	 */
103 104
	tick_length		+= new_base - tick_length_base;
	tick_length_base	 = new_base;
105 106
}

107
static inline s64 ntp_update_offset_fll(s64 offset64, long secs)
108 109 110 111
{
	time_status &= ~STA_MODE;

	if (secs < MINSEC)
112
		return 0;
113 114

	if (!(time_status & STA_FLL) && (secs <= MAXSEC))
115
		return 0;
116 117 118

	time_status |= STA_MODE;

119
	return div_s64(offset64 << (NTP_SCALE_SHIFT - SHIFT_FLL), secs);
120 121
}

R
Roman Zippel 已提交
122 123 124
static void ntp_update_offset(long offset)
{
	s64 freq_adj;
125 126
	s64 offset64;
	long secs;
R
Roman Zippel 已提交
127 128 129 130

	if (!(time_status & STA_PLL))
		return;

R
Roman Zippel 已提交
131
	if (!(time_status & STA_NANO))
132
		offset *= NSEC_PER_USEC;
R
Roman Zippel 已提交
133 134 135 136 137

	/*
	 * Scale the phase adjustment and
	 * clamp to the operating range.
	 */
138 139
	offset = min(offset, MAXPHASE);
	offset = max(offset, -MAXPHASE);
R
Roman Zippel 已提交
140 141 142 143 144

	/*
	 * Select how the frequency is to be controlled
	 * and in which mode (PLL or FLL).
	 */
145
	secs = xtime.tv_sec - time_reftime;
146
	if (unlikely(time_status & STA_FREQHOLD))
147 148
		secs = 0;

R
Roman Zippel 已提交
149 150
	time_reftime = xtime.tv_sec;

151 152 153 154
	offset64    = offset;
	freq_adj    = (offset64 * secs) <<
			(NTP_SCALE_SHIFT - 2 * (SHIFT_PLL + 2 + time_constant));

155
	freq_adj    += ntp_update_offset_fll(offset64, secs);
156 157 158 159 160 161

	freq_adj    = min(freq_adj + time_freq, MAXFREQ_SCALED);

	time_freq   = max(freq_adj, -MAXFREQ_SCALED);

	time_offset = div_s64(offset64 << NTP_SCALE_SHIFT, NTP_INTERVAL_FREQ);
R
Roman Zippel 已提交
162 163
}

164 165 166 167 168 169 170
/**
 * ntp_clear - Clears the NTP state variables
 *
 * Must be called while holding a write on the xtime_lock
 */
void ntp_clear(void)
{
171 172 173 174
	time_adjust	= 0;		/* stop active adjtime() */
	time_status	|= STA_UNSYNC;
	time_maxerror	= NTP_PHASE_LIMIT;
	time_esterror	= NTP_PHASE_LIMIT;
175 176 177

	ntp_update_frequency();

178 179
	tick_length	= tick_length_base;
	time_offset	= 0;
180 181
}

182
/*
R
Roman Zippel 已提交
183 184 185
 * Leap second processing. If in leap-insert state at the end of the
 * day, the system clock is set back one second; if in leap-delete
 * state, the system clock is set ahead one second.
186
 */
R
Roman Zippel 已提交
187
static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer)
188
{
R
Roman Zippel 已提交
189
	enum hrtimer_restart res = HRTIMER_NORESTART;
190

191
	write_seqlock(&xtime_lock);
192 193 194 195 196

	switch (time_state) {
	case TIME_OK:
		break;
	case TIME_INS:
R
Roman Zippel 已提交
197 198 199
		xtime.tv_sec--;
		wall_to_monotonic.tv_sec++;
		time_state = TIME_OOP;
200 201
		printk(KERN_NOTICE
			"Clock: inserting leap second 23:59:60 UTC\n");
202
		hrtimer_add_expires_ns(&leap_timer, NSEC_PER_SEC);
R
Roman Zippel 已提交
203
		res = HRTIMER_RESTART;
204 205
		break;
	case TIME_DEL:
R
Roman Zippel 已提交
206 207 208 209
		xtime.tv_sec++;
		time_tai--;
		wall_to_monotonic.tv_sec--;
		time_state = TIME_WAIT;
210 211
		printk(KERN_NOTICE
			"Clock: deleting leap second 23:59:59 UTC\n");
212 213
		break;
	case TIME_OOP:
R
Roman Zippel 已提交
214
		time_tai++;
215
		time_state = TIME_WAIT;
R
Roman Zippel 已提交
216
		/* fall through */
217 218
	case TIME_WAIT:
		if (!(time_status & (STA_INS | STA_DEL)))
R
Roman Zippel 已提交
219
			time_state = TIME_OK;
R
Roman Zippel 已提交
220 221 222 223
		break;
	}
	update_vsyscall(&xtime, clock);

224
	write_sequnlock(&xtime_lock);
R
Roman Zippel 已提交
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245

	return res;
}

/*
 * this routine handles the overflow of the microsecond field
 *
 * The tricky bits of code to handle the accurate clock support
 * were provided by Dave Mills (Mills@UDEL.EDU) of NTP fame.
 * They were originally developed for SUN and DEC kernels.
 * All the kudos should go to Dave for this stuff.
 */
void second_overflow(void)
{
	s64 time_adj;

	/* Bump the maxerror field */
	time_maxerror += MAXFREQ / NSEC_PER_USEC;
	if (time_maxerror > NTP_PHASE_LIMIT) {
		time_maxerror = NTP_PHASE_LIMIT;
		time_status |= STA_UNSYNC;
246 247 248
	}

	/*
249 250
	 * Compute the phase adjustment for the next second. The offset is
	 * reduced by a fixed factor times the time constant.
251
	 */
252 253 254 255
	tick_length	= tick_length_base;
	time_adj	= shift_right(time_offset, SHIFT_PLL + time_constant);
	time_offset	-= time_adj;
	tick_length	+= time_adj;
256

257 258 259 260 261 262 263
	if (!time_adjust)
		return;

	if (time_adjust > MAX_TICKADJ) {
		time_adjust -= MAX_TICKADJ;
		tick_length += MAX_TICKADJ_SCALED;
		return;
264
	}
265 266 267 268 269 270 271 272 273 274

	if (time_adjust < -MAX_TICKADJ) {
		time_adjust += MAX_TICKADJ;
		tick_length -= MAX_TICKADJ_SCALED;
		return;
	}

	tick_length += (s64)(time_adjust * NSEC_PER_USEC / NTP_INTERVAL_FREQ)
							 << NTP_SCALE_SHIFT;
	time_adjust = 0;
275 276
}

277
#ifdef CONFIG_GENERIC_CMOS_UPDATE
278

279 280 281
/* Disable the cmos update - used by virtualization and embedded */
int no_sync_cmos_clock  __read_mostly;

282
static void sync_cmos_clock(struct work_struct *work);
283

284
static DECLARE_DELAYED_WORK(sync_cmos_work, sync_cmos_clock);
285

286
static void sync_cmos_clock(struct work_struct *work)
287 288 289 290 291 292 293 294 295 296 297
{
	struct timespec now, next;
	int fail = 1;

	/*
	 * If we have an externally synchronized Linux clock, then update
	 * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be
	 * called as close as possible to 500 ms before the new second starts.
	 * This code is run on a timer.  If the clock is set, that timer
	 * may not expire at the correct time.  Thus, we adjust...
	 */
298
	if (!ntp_synced()) {
299 300 301 302 303
		/*
		 * Not synced, exit, do not restart a timer (if one is
		 * running, let it run out).
		 */
		return;
304
	}
305 306

	getnstimeofday(&now);
307
	if (abs(now.tv_nsec - (NSEC_PER_SEC / 2)) <= tick_nsec / 2)
308 309
		fail = update_persistent_clock(now);

310
	next.tv_nsec = (NSEC_PER_SEC / 2) - now.tv_nsec - (TICK_NSEC / 2);
311 312 313 314 315 316 317 318 319 320 321 322
	if (next.tv_nsec <= 0)
		next.tv_nsec += NSEC_PER_SEC;

	if (!fail)
		next.tv_sec = 659;
	else
		next.tv_sec = 0;

	if (next.tv_nsec >= NSEC_PER_SEC) {
		next.tv_sec++;
		next.tv_nsec -= NSEC_PER_SEC;
	}
323
	schedule_delayed_work(&sync_cmos_work, timespec_to_jiffies(&next));
324 325 326
}

static void notify_cmos_timer(void)
327
{
328
	if (!no_sync_cmos_clock)
329
		schedule_delayed_work(&sync_cmos_work, 0);
330 331
}

332 333 334 335
#else
static inline void notify_cmos_timer(void) { }
#endif

336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356
/*
 * Start the leap seconds timer:
 */
static inline void ntp_start_leap_timer(struct timespec *ts)
{
	long now = ts->tv_sec;

	if (time_status & STA_INS) {
		time_state = TIME_INS;
		now += 86400 - now % 86400;
		hrtimer_start(&leap_timer, ktime_set(now, 0), HRTIMER_MODE_ABS);

		return;
	}

	if (time_status & STA_DEL) {
		time_state = TIME_DEL;
		now += 86400 - (now + 1) % 86400;
		hrtimer_start(&leap_timer, ktime_set(now, 0), HRTIMER_MODE_ABS);
	}
}
I
Ingo Molnar 已提交
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380

/*
 * Propagate a new txc->status value into the NTP state:
 */
static inline void process_adj_status(struct timex *txc, struct timespec *ts)
{
	if ((time_status & STA_PLL) && !(txc->status & STA_PLL)) {
		time_state = TIME_OK;
		time_status = STA_UNSYNC;
	}
	/* only set allowed bits */
	time_status &= STA_RONLY;

	/*
	 * If we turn on PLL adjustments then reset the
	 * reference time to current time.
	 */
	if (!(time_status & STA_PLL) && (txc->status & STA_PLL))
		time_reftime = xtime.tv_sec;

	time_status |= txc->status & ~STA_RONLY;

	switch (time_state) {
	case TIME_OK:
381
		ntp_start_leap_timer(ts);
I
Ingo Molnar 已提交
382 383 384 385
		break;
	case TIME_INS:
	case TIME_DEL:
		time_state = TIME_OK;
386
		ntp_start_leap_timer(ts);
I
Ingo Molnar 已提交
387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
	case TIME_WAIT:
		if (!(time_status & (STA_INS | STA_DEL)))
			time_state = TIME_OK;
		break;
	case TIME_OOP:
		hrtimer_restart(&leap_timer);
		break;
	}
}
/*
 * Called with the xtime lock held, so we can access and modify
 * all the global NTP state:
 */
static inline void process_adjtimex_modes(struct timex *txc, struct timespec *ts)
{
	if (txc->modes & ADJ_STATUS)
		process_adj_status(txc, ts);

	if (txc->modes & ADJ_NANO)
		time_status |= STA_NANO;
407

I
Ingo Molnar 已提交
408 409 410 411
	if (txc->modes & ADJ_MICRO)
		time_status &= ~STA_NANO;

	if (txc->modes & ADJ_FREQUENCY) {
412
		time_freq = txc->freq * PPM_SCALE;
I
Ingo Molnar 已提交
413 414 415 416 417 418
		time_freq = min(time_freq, MAXFREQ_SCALED);
		time_freq = max(time_freq, -MAXFREQ_SCALED);
	}

	if (txc->modes & ADJ_MAXERROR)
		time_maxerror = txc->maxerror;
419

I
Ingo Molnar 已提交
420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
	if (txc->modes & ADJ_ESTERROR)
		time_esterror = txc->esterror;

	if (txc->modes & ADJ_TIMECONST) {
		time_constant = txc->constant;
		if (!(time_status & STA_NANO))
			time_constant += 4;
		time_constant = min(time_constant, (long)MAXTC);
		time_constant = max(time_constant, 0l);
	}

	if (txc->modes & ADJ_TAI && txc->constant > 0)
		time_tai = txc->constant;

	if (txc->modes & ADJ_OFFSET)
		ntp_update_offset(txc->offset);
436

I
Ingo Molnar 已提交
437 438 439 440 441 442 443
	if (txc->modes & ADJ_TICK)
		tick_usec = txc->tick;

	if (txc->modes & (ADJ_TICK|ADJ_FREQUENCY|ADJ_OFFSET))
		ntp_update_frequency();
}

444 445
/*
 * adjtimex mainly allows reading (and writing, if superuser) of
446 447 448 449
 * kernel time-keeping variables. used by xntpd.
 */
int do_adjtimex(struct timex *txc)
{
R
Roman Zippel 已提交
450
	struct timespec ts;
451 452
	int result;

453 454
	/* Validate the data before disabling interrupts */
	if (txc->modes & ADJ_ADJTIME) {
R
Roman Zippel 已提交
455
		/* singleshot must not be used with any other mode bits */
456
		if (!(txc->modes & ADJ_OFFSET_SINGLESHOT))
457
			return -EINVAL;
458 459 460 461 462 463 464 465
		if (!(txc->modes & ADJ_OFFSET_READONLY) &&
		    !capable(CAP_SYS_TIME))
			return -EPERM;
	} else {
		/* In order to modify anything, you gotta be super-user! */
		 if (txc->modes && !capable(CAP_SYS_TIME))
			return -EPERM;

466 467 468 469
		/*
		 * if the quartz is off by more than 10% then
		 * something is VERY wrong!
		 */
470 471 472
		if (txc->modes & ADJ_TICK &&
		    (txc->tick <  900000/USER_HZ ||
		     txc->tick > 1100000/USER_HZ))
473
			return -EINVAL;
474 475 476

		if (txc->modes & ADJ_STATUS && time_state != TIME_OK)
			hrtimer_cancel(&leap_timer);
J
John Stultz 已提交
477
	}
478

R
Roman Zippel 已提交
479 480
	getnstimeofday(&ts);

481 482
	write_seqlock_irq(&xtime_lock);

483 484 485 486 487 488 489 490 491
	if (txc->modes & ADJ_ADJTIME) {
		long save_adjust = time_adjust;

		if (!(txc->modes & ADJ_OFFSET_READONLY)) {
			/* adjtime() is independent from ntp_adjtime() */
			time_adjust = txc->offset;
			ntp_update_frequency();
		}
		txc->offset = save_adjust;
492
	} else {
R
Roman Zippel 已提交
493

494 495 496
		/* If there are input parameters, then process them: */
		if (txc->modes)
			process_adjtimex_modes(txc, &ts);
R
Roman Zippel 已提交
497

498
		txc->offset = shift_right(time_offset * NTP_INTERVAL_FREQ,
499
				  NTP_SCALE_SHIFT);
500 501 502
		if (!(time_status & STA_NANO))
			txc->offset /= NSEC_PER_USEC;
	}
503

R
Roman Zippel 已提交
504
	result = time_state;	/* mostly `TIME_OK' */
R
Roman Zippel 已提交
505
	if (time_status & (STA_UNSYNC|STA_CLOCKERR))
506 507
		result = TIME_ERROR;

508
	txc->freq	   = shift_right((time_freq >> PPM_SCALE_INV_SHIFT) *
509
					 PPM_SCALE_INV, NTP_SCALE_SHIFT);
510 511 512 513
	txc->maxerror	   = time_maxerror;
	txc->esterror	   = time_esterror;
	txc->status	   = time_status;
	txc->constant	   = time_constant;
514
	txc->precision	   = 1;
515
	txc->tolerance	   = MAXFREQ_SCALED / PPM_SCALE;
516
	txc->tick	   = tick_usec;
R
Roman Zippel 已提交
517
	txc->tai	   = time_tai;
518 519 520 521 522 523 524 525 526 527

	/* PPS is not implemented, so these are zero */
	txc->ppsfreq	   = 0;
	txc->jitter	   = 0;
	txc->shift	   = 0;
	txc->stabil	   = 0;
	txc->jitcnt	   = 0;
	txc->calcnt	   = 0;
	txc->errcnt	   = 0;
	txc->stbcnt	   = 0;
528

529
	write_sequnlock_irq(&xtime_lock);
R
Roman Zippel 已提交
530

R
Roman Zippel 已提交
531 532 533 534
	txc->time.tv_sec = ts.tv_sec;
	txc->time.tv_usec = ts.tv_nsec;
	if (!(time_status & STA_NANO))
		txc->time.tv_usec /= NSEC_PER_USEC;
R
Roman Zippel 已提交
535

536
	notify_cmos_timer();
R
Roman Zippel 已提交
537 538

	return result;
539
}
540 541 542 543

static int __init ntp_tick_adj_setup(char *str)
{
	ntp_tick_adj = simple_strtol(str, NULL, 0);
544 545
	ntp_tick_adj <<= NTP_SCALE_SHIFT;

546 547 548 549
	return 1;
}

__setup("ntp_tick_adj=", ntp_tick_adj_setup);
R
Roman Zippel 已提交
550 551 552 553 554 555 556

void __init ntp_init(void)
{
	ntp_clear();
	hrtimer_init(&leap_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	leap_timer.function = ntp_leap_second;
}