summaryrefslogtreecommitdiffstats
path: root/arch/x86/xen/time.c
blob: 1d597364b49dc3f9b7e8d259f1300181d4beab8e (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
// SPDX-License-Identifier: GPL-2.0
/*
 * Xen time implementation.
 *
 * This is implemented in terms of a clocksource driver which uses
 * the hypervisor clock as a nanosecond timebase, and a clockevent
 * driver which uses the hypervisor's timer mechanism.
 *
 * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
 */
#include <linux/kernel.h>
#include <linux/interrupt.h>
#include <linux/clocksource.h>
#include <linux/clockchips.h>
#include <linux/gfp.h>
#include <linux/slab.h>
#include <linux/pvclock_gtod.h>
#include <linux/timekeeper_internal.h>

#include <asm/pvclock.h>
#include <asm/xen/hypervisor.h>
#include <asm/xen/hypercall.h>

#include <xen/events.h>
#include <xen/features.h>
#include <xen/interface/xen.h>
#include <xen/interface/vcpu.h>

#include "xen-ops.h"

/* Minimum amount of time until next clock event fires */
#define TIMER_SLOP	100000

static u64 xen_sched_clock_offset __read_mostly;

/* Get the TSC speed from Xen */
static unsigned long xen_tsc_khz(void)
{
	struct pvclock_vcpu_time_info *info =
		&HYPERVISOR_shared_info->vcpu_info[0].time;

	setup_force_cpu_cap(X86_FEATURE_TSC_KNOWN_FREQ);
	return pvclock_tsc_khz(info);
}

static u64 xen_clocksource_read(void)
{
        struct pvclock_vcpu_time_info *src;
	u64 ret;

	preempt_disable_notrace();
	src = &__this_cpu_read(xen_vcpu)->time;
	ret = pvclock_clocksource_read(src);
	preempt_enable_notrace();
	return ret;
}

static u64 xen_clocksource_get_cycles(struct clocksource *cs)
{
	return xen_clocksource_read();
}

static noinstr u64 xen_sched_clock(void)
{
        struct pvclock_vcpu_time_info *src;
	u64 ret;

	preempt_disable_notrace();
	src = &__this_cpu_read(xen_vcpu)->time;
	ret = pvclock_clocksource_read_nowd(src);
	ret -= xen_sched_clock_offset;
	preempt_enable_notrace();
	return ret;
}

static void xen_read_wallclock(struct timespec64 *ts)
{
	struct shared_info *s = HYPERVISOR_shared_info;
	struct pvclock_wall_clock *wall_clock = &(s->wc);
        struct pvclock_vcpu_time_info *vcpu_time;

	vcpu_time = &get_cpu_var(xen_vcpu)->time;
	pvclock_read_wallclock(wall_clock, vcpu_time, ts);
	put_cpu_var(xen_vcpu);
}

static void xen_get_wallclock(struct timespec64 *now)
{
	xen_read_wallclock(now);
}

static int xen_set_wallclock(const struct timespec64 *now)
{
	return -ENODEV;
}

static int xen_pvclock_gtod_notify(struct notifier_block *nb,
				   unsigned long was_set, void *priv)
{
	/* Protected by the calling core code serialization */
	static struct timespec64 next_sync;

	struct xen_platform_op op;
	struct timespec64 now;
	struct timekeeper *tk = priv;
	static bool settime64_supported = true;
	int ret;

	now.tv_sec = tk->xtime_sec;
	now.tv_nsec = (long)(tk->tkr_mono.xtime_nsec >> tk->tkr_mono.shift);

	/*
	 * We only take the expensive HV call when the clock was set
	 * or when the 11 minutes RTC synchronization time elapsed.
	 */
	if (!was_set && timespec64_compare(&now, &next_sync) < 0)
		return NOTIFY_OK;

again:
	if (settime64_supported) {
		op.cmd = XENPF_settime64;
		op.u.settime64.mbz = 0;
		op.u.settime64.secs = now.tv_sec;
		op.u.settime64.nsecs = now.tv_nsec;
		op.u.settime64.system_time = xen_clocksource_read();
	} else {
		op.cmd = XENPF_settime32;
		op.u.settime32.secs = now.tv_sec;
		op.u.settime32.nsecs = now.tv_nsec;
		op.u.settime32.system_time = xen_clocksource_read();
	}

	ret = HYPERVISOR_platform_op(&op);

	if (ret == -ENOSYS && settime64_supported) {
		settime64_supported = false;
		goto again;
	}
	if (ret < 0)
		return NOTIFY_BAD;

	/*
	 * Move the next drift compensation time 11 minutes
	 * ahead. That's emulating the sync_cmos_clock() update for
	 * the hardware RTC.
	 */
	next_sync = now;
	next_sync.tv_sec += 11 * 60;

	return NOTIFY_OK;
}

static struct notifier_block xen_pvclock_gtod_notifier = {
	.notifier_call = xen_pvclock_gtod_notify,
};

static int xen_cs_enable(struct clocksource *cs)
{
	vclocks_set_used(VDSO_CLOCKMODE_PVCLOCK);
	return 0;
}

static struct clocksource xen_clocksource __read_mostly = {
	.name	= "xen",
	.rating	= 400,
	.read	= xen_clocksource_get_cycles,
	.mask	= CLOCKSOURCE_MASK(64),
	.flags	= CLOCK_SOURCE_IS_CONTINUOUS,
	.enable = xen_cs_enable,
};

/*
   Xen clockevent implementation

   Xen has two clockevent implementations:

   The old timer_op one works with all released versions of Xen prior
   to version 3.0.4.  This version of the hypervisor provides a
   single-shot timer with nanosecond resolution.  However, sharing the
   same event channel is a 100Hz tick which is delivered while the
   vcpu is running.  We don't care about or use this tick, but it will
   cause the core time code to think the timer fired too soon, and
   will end up resetting it each time.  It could be filtered, but
   doing so has complications when the ktime clocksource is not yet
   the xen clocksource (ie, at boot time).

   The new vcpu_op-based timer interface allows the tick timer period
   to be changed or turned off.  The tick timer is not useful as a
   periodic timer because events are only delivered to running vcpus.
   The one-shot timer can report when a timeout is in the past, so
   set_next_event is capable of returning -ETIME when appropriate.
   This interface is used when available.
*/


/*
  Get a hypervisor absolute time.  In theory we could maintain an
  offset between the kernel's time and the hypervisor's time, and
  apply that to a kernel's absolute timeout.  Unfortunately the
  hypervisor and kernel times can drift even if the kernel is using
  the Xen clocksource, because ntp can warp the kernel's clocksource.
*/
static s64 get_abs_timeout(unsigned long delta)
{
	return xen_clocksource_read() + delta;
}

static int xen_timerop_shutdown(struct clock_event_device *evt)
{
	/* cancel timeout */
	HYPERVISOR_set_timer_op(0);

	return 0;
}

static int xen_timerop_set_next_event(unsigned long delta,
				      struct clock_event_device *evt)
{
	WARN_ON(!clockevent_state_oneshot(evt));

	if (HYPERVISOR_set_timer_op(get_abs_timeout(delta)) < 0)
		BUG();

	/* We may have missed the deadline, but there's no real way of
	   knowing for sure.  If the event was in the past, then we'll
	   get an immediate interrupt. */

	return 0;
}

static struct clock_event_device xen_timerop_clockevent __ro_after_init = {
	.name			= "xen",
	.features		= CLOCK_EVT_FEAT_ONESHOT,

	.max_delta_ns		= 0xffffffff,
	.max_delta_ticks	= 0xffffffff,
	.min_delta_ns		= TIMER_SLOP,
	.min_delta_ticks	= TIMER_SLOP,

	.mult			= 1,
	.shift			= 0,
	.rating			= 500,

	.set_state_shutdown	= xen_timerop_shutdown,
	.set_next_event		= xen_timerop_set_next_event,
};

static int xen_vcpuop_shutdown(struct clock_event_device *evt)
{
	int cpu = smp_processor_id();

	if (HYPERVISOR_vcpu_op(VCPUOP_stop_singleshot_timer, xen_vcpu_nr(cpu),
			       NULL) ||
	    HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, xen_vcpu_nr(cpu),
			       NULL))
		BUG();

	return 0;
}

static int xen_vcpuop_set_oneshot(struct clock_event_device *evt)
{
	int cpu = smp_processor_id();

	if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, xen_vcpu_nr(cpu),
			       NULL))
		BUG();

	return 0;
}

static int xen_vcpuop_set_next_event(unsigned long delta,
				     struct clock_event_device *evt)
{
	int cpu = smp_processor_id();
	struct vcpu_set_singleshot_timer single;
	int ret;

	WARN_ON(!clockevent_state_oneshot(evt));

	single.timeout_abs_ns = get_abs_timeout(delta);
	/* Get an event anyway, even if the timeout is already expired */
	single.flags = 0;

	ret = HYPERVISOR_vcpu_op(VCPUOP_set_singleshot_timer, xen_vcpu_nr(cpu),
				 &single);
	BUG_ON(ret != 0);

	return ret;
}

static struct clock_event_device xen_vcpuop_clockevent __ro_after_init = {
	.name = "xen",
	.features = CLOCK_EVT_FEAT_ONESHOT,

	.max_delta_ns = 0xffffffff,
	.max_delta_ticks = 0xffffffff,
	.min_delta_ns = TIMER_SLOP,
	.min_delta_ticks = TIMER_SLOP,

	.mult = 1,
	.shift = 0,
	.rating = 500,

	.set_state_shutdown = xen_vcpuop_shutdown,
	.set_state_oneshot = xen_vcpuop_set_oneshot,
	.set_next_event = xen_vcpuop_set_next_event,
};

static const struct clock_event_device *xen_clockevent =
	&xen_timerop_clockevent;

struct xen_clock_event_device {
	struct clock_event_device evt;
	char name[16];
};
static DEFINE_PER_CPU(struct xen_clock_event_device, xen_clock_events) = { .evt.irq = -1 };

static irqreturn_t xen_timer_interrupt(int irq, void *dev_id)
{
	struct clock_event_device *evt = this_cpu_ptr(&xen_clock_events.evt);
	irqreturn_t ret;

	ret = IRQ_NONE;
	if (evt->event_handler) {
		evt->event_handler(evt);
		ret = IRQ_HANDLED;
	}

	return ret;
}

void xen_teardown_timer(int cpu)
{
	struct clock_event_device *evt;
	evt = &per_cpu(xen_clock_events, cpu).evt;

	if (evt->irq >= 0) {
		unbind_from_irqhandler(evt->irq, NULL);
		evt->irq = -1;
	}
}

void xen_setup_timer(int cpu)
{
	struct xen_clock_event_device *xevt = &per_cpu(xen_clock_events, cpu);
	struct clock_event_device *evt = &xevt->evt;
	int irq;

	WARN(evt->irq >= 0, "IRQ%d for CPU%d is already allocated\n", evt->irq, cpu);
	if (evt->irq >= 0)
		xen_teardown_timer(cpu);

	printk(KERN_INFO "installing Xen timer for CPU %d\n", cpu);

	snprintf(xevt->name, sizeof(xevt->name), "timer%d", cpu);

	irq = bind_virq_to_irqhandler(VIRQ_TIMER, cpu, xen_timer_interrupt,
				      IRQF_PERCPU|IRQF_NOBALANCING|IRQF_TIMER|
				      IRQF_FORCE_RESUME|IRQF_EARLY_RESUME,
				      xevt->name, NULL);
	(void)xen_set_irq_priority(irq, XEN_IRQ_PRIORITY_MAX);

	memcpy(evt, xen_clockevent, sizeof(*evt));

	evt->cpumask = cpumask_of(cpu);
	evt->irq = irq;
}


void xen_setup_cpu_clockevents(void)
{
	clockevents_register_device(this_cpu_ptr(&xen_clock_events.evt));
}

void xen_timer_resume(void)
{
	int cpu;

	if (xen_clockevent != &xen_vcpuop_clockevent)
		return;

	for_each_online_cpu(cpu) {
		if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer,
				       xen_vcpu_nr(cpu), NULL))
			BUG();
	}
}

static struct pvclock_vsyscall_time_info *xen_clock __read_mostly;
static u64 xen_clock_value_saved;

void xen_save_time_memory_area(void)
{
	struct vcpu_register_time_memory_area t;
	int ret;

	xen_clock_value_saved = xen_clocksource_read() - xen_sched_clock_offset;

	if (!xen_clock)
		return;

	t.addr.v = NULL;

	ret = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_time_memory_area, 0, &t);
	if (ret != 0)
		pr_notice("Cannot save secondary vcpu_time_info (err %d)",
			  ret);
	else
		clear_page(xen_clock);
}

void xen_restore_time_memory_area(void)
{
	struct vcpu_register_time_memory_area t;
	int ret;

	if (!xen_clock)
		goto out;

	t.addr.v = &xen_clock->pvti;

	ret = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_time_memory_area, 0, &t);

	/*
	 * We don't disable VDSO_CLOCKMODE_PVCLOCK entirely if it fails to
	 * register the secondary time info with Xen or if we migrated to a
	 * host without the necessary flags. On both of these cases what
	 * happens is either process seeing a zeroed out pvti or seeing no
	 * PVCLOCK_TSC_STABLE_BIT bit set. Userspace checks the latter and
	 * if 0, it discards the data in pvti and fallbacks to a system
	 * call for a reliable timestamp.
	 */
	if (ret != 0)
		pr_notice("Cannot restore secondary vcpu_time_info (err %d)",
			  ret);

out:
	/* Need pvclock_resume() before using xen_clocksource_read(). */
	pvclock_resume();
	xen_sched_clock_offset = xen_clocksource_read() - xen_clock_value_saved;
}

static void xen_setup_vsyscall_time_info(void)
{
	struct vcpu_register_time_memory_area t;
	struct pvclock_vsyscall_time_info *ti;
	int ret;

	ti = (struct pvclock_vsyscall_time_info *)get_zeroed_page(GFP_KERNEL);
	if (!ti)
		return;

	t.addr.v = &ti->pvti;

	ret = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_time_memory_area, 0, &t);
	if (ret) {
		pr_notice("xen: VDSO_CLOCKMODE_PVCLOCK not supported (err %d)\n", ret);
		free_page((unsigned long)ti);
		return;
	}

	/*
	 * If primary time info had this bit set, secondary should too since
	 * it's the same data on both just different memory regions. But we
	 * still check it in case hypervisor is buggy.
	 */
	if (!(ti->pvti.flags & PVCLOCK_TSC_STABLE_BIT)) {
		t.addr.v = NULL;
		ret = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_time_memory_area,
					 0, &t);
		if (!ret)
			free_page((unsigned long)ti);

		pr_notice("xen: VDSO_CLOCKMODE_PVCLOCK not supported (tsc unstable)\n");
		return;
	}

	xen_clock = ti;
	pvclock_set_pvti_cpu0_va(xen_clock);

	xen_clocksource.vdso_clock_mode = VDSO_CLOCKMODE_PVCLOCK;
}

/*
 * Check if it is possible to safely use the tsc as a clocksource.  This is
 * only true if the hypervisor notifies the guest that its tsc is invariant,
 * the tsc is stable, and the tsc instruction will never be emulated.
 */
static int __init xen_tsc_safe_clocksource(void)
{
	u32 eax, ebx, ecx, edx;

	if (!(boot_cpu_has(X86_FEATURE_CONSTANT_TSC)))
		return 0;

	if (!(boot_cpu_has(X86_FEATURE_NONSTOP_TSC)))
		return 0;

	if (check_tsc_unstable())
		return 0;

	/* Leaf 4, sub-leaf 0 (0x40000x03) */
	cpuid_count(xen_cpuid_base() + 3, 0, &eax, &ebx, &ecx, &edx);

	/* tsc_mode = no_emulate (2) */
	if (ebx != 2)
		return 0;

	return 1;
}

static void __init xen_time_init(void)
{
	struct pvclock_vcpu_time_info *pvti;
	int cpu = smp_processor_id();
	struct timespec64 tp;

	/*
	 * As Dom0 is never moved, no penalty on using TSC there.
	 *
	 * If it is possible for the guest to determine that the tsc is a safe
	 * clocksource, then set xen_clocksource rating below that of the tsc
	 * so that the system prefers tsc instead.
	 */
	if (xen_initial_domain())
		xen_clocksource.rating = 275;
	else if (xen_tsc_safe_clocksource())
		xen_clocksource.rating = 299;

	clocksource_register_hz(&xen_clocksource, NSEC_PER_SEC);

	if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, xen_vcpu_nr(cpu),
			       NULL) == 0) {
		/* Successfully turned off 100Hz tick, so we have the
		   vcpuop-based timer interface */
		printk(KERN_DEBUG "Xen: using vcpuop timer interface\n");
		xen_clockevent = &xen_vcpuop_clockevent;
	}

	/* Set initial system time with full resolution */
	xen_read_wallclock(&tp);
	do_settimeofday64(&tp);

	setup_force_cpu_cap(X86_FEATURE_TSC);

	/*
	 * We check ahead on the primary time info if this
	 * bit is supported hence speeding up Xen clocksource.
	 */
	pvti = &__this_cpu_read(xen_vcpu)->time;
	if (pvti->flags & PVCLOCK_TSC_STABLE_BIT) {
		pvclock_set_flags(PVCLOCK_TSC_STABLE_BIT);
		xen_setup_vsyscall_time_info();
	}

	xen_setup_runstate_info(cpu);
	xen_setup_timer(cpu);
	xen_setup_cpu_clockevents();

	xen_time_setup_guest();

	if (xen_initial_domain())
		pvclock_gtod_register_notifier(&xen_pvclock_gtod_notifier);
}

static void __init xen_init_time_common(void)
{
	xen_sched_clock_offset = xen_clocksource_read();
	static_call_update(pv_steal_clock, xen_steal_clock);
	paravirt_set_sched_clock(xen_sched_clock);

	x86_platform.calibrate_tsc = xen_tsc_khz;
	x86_platform.get_wallclock = xen_get_wallclock;
}

void __init xen_init_time_ops(void)
{
	xen_init_time_common();

	x86_init.timers.timer_init = xen_time_init;
	x86_init.timers.setup_percpu_clockev = x86_init_noop;
	x86_cpuinit.setup_percpu_clockev = x86_init_noop;

	/* Dom0 uses the native method to set the hardware RTC. */
	if (!xen_initial_domain())
		x86_platform.set_wallclock = xen_set_wallclock;
}

#ifdef CONFIG_XEN_PVHVM
static void xen_hvm_setup_cpu_clockevents(void)
{
	int cpu = smp_processor_id();
	xen_setup_runstate_info(cpu);
	/*
	 * xen_setup_timer(cpu) - snprintf is bad in atomic context. Hence
	 * doing it xen_hvm_cpu_notify (which gets called by smp_init during
	 * early bootup and also during CPU hotplug events).
	 */
	xen_setup_cpu_clockevents();
}

void __init xen_hvm_init_time_ops(void)
{
	static bool hvm_time_initialized;

	if (hvm_time_initialized)
		return;

	/*
	 * vector callback is needed otherwise we cannot receive interrupts
	 * on cpu > 0 and at this point we don't know how many cpus are
	 * available.
	 */
	if (!xen_have_vector_callback)
		return;

	if (!xen_feature(XENFEAT_hvm_safe_pvclock)) {
		pr_info_once("Xen doesn't support pvclock on HVM, disable pv timer");
		return;
	}

	/*
	 * Only MAX_VIRT_CPUS 'vcpu_info' are embedded inside 'shared_info'.
	 * The __this_cpu_read(xen_vcpu) is still NULL when Xen HVM guest
	 * boots on vcpu >= MAX_VIRT_CPUS (e.g., kexec), To access
	 * __this_cpu_read(xen_vcpu) via xen_clocksource_read() will panic.
	 *
	 * The xen_hvm_init_time_ops() should be called again later after
	 * __this_cpu_read(xen_vcpu) is available.
	 */
	if (!__this_cpu_read(xen_vcpu)) {
		pr_info("Delay xen_init_time_common() as kernel is running on vcpu=%d\n",
			xen_vcpu_nr(0));
		return;
	}

	xen_init_time_common();

	x86_init.timers.setup_percpu_clockev = xen_time_init;
	x86_cpuinit.setup_percpu_clockev = xen_hvm_setup_cpu_clockevents;

	x86_platform.set_wallclock = xen_set_wallclock;

	hvm_time_initialized = true;
}
#endif

/* Kernel parameter to specify Xen timer slop */
static int __init parse_xen_timer_slop(char *ptr)
{
	unsigned long slop = memparse(ptr, NULL);

	xen_timerop_clockevent.min_delta_ns = slop;
	xen_timerop_clockevent.min_delta_ticks = slop;
	xen_vcpuop_clockevent.min_delta_ns = slop;
	xen_vcpuop_clockevent.min_delta_ticks = slop;

	return 0;
}
early_param("xen_timer_slop", parse_xen_timer_slop);