kvmclock.c 4.59 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
/*  KVM paravirtual clock driver. A clocksource implementation
    Copyright (C) 2008 Glauber de Oliveira Costa, Red Hat Inc.

    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program; if not, write to the Free Software
    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
*/

#include <linux/clocksource.h>
#include <linux/kvm_para.h>
21
#include <asm/pvclock.h>
22
23
24
25
#include <asm/arch_hooks.h>
#include <asm/msr.h>
#include <asm/apic.h>
#include <linux/percpu.h>
26
#include <asm/reboot.h>
27
28
29
30
31
32
33
34
35
36
37
38
39

#define KVM_SCALE 22

static int kvmclock = 1;

static int parse_no_kvmclock(char *arg)
{
	kvmclock = 0;
	return 0;
}
early_param("no-kvmclock", parse_no_kvmclock);

/* The hypervisor will put information about time periodically here */
40
41
static DEFINE_PER_CPU_SHARED_ALIGNED(struct pvclock_vcpu_time_info, hv_clock);
static struct pvclock_wall_clock wall_clock;
42
43
44
45
46
47

/*
 * The wallclock is the time of day when we booted. Since then, some time may
 * have elapsed since the hypervisor wrote the data. So we try to account for
 * that with system time
 */
Ingo Molnar's avatar
Ingo Molnar committed
48
static unsigned long kvm_get_wallclock(void)
49
{
50
	struct pvclock_vcpu_time_info *vcpu_time;
51
52
53
54
55
	struct timespec ts;
	int low, high;

	low = (int)__pa(&wall_clock);
	high = ((u64)__pa(&wall_clock) >> 32);
56
	native_write_msr(MSR_KVM_WALL_CLOCK, low, high);
57

58
59
60
	vcpu_time = &get_cpu_var(hv_clock);
	pvclock_read_wallclock(&wall_clock, vcpu_time, &ts);
	put_cpu_var(hv_clock);
61

62
	return ts.tv_sec;
63
64
}

Ingo Molnar's avatar
Ingo Molnar committed
65
static int kvm_set_wallclock(unsigned long now)
66
{
67
	return -1;
68
69
70
71
}

static cycle_t kvm_clock_read(void)
{
72
73
	struct pvclock_vcpu_time_info *src;
	cycle_t ret;
74

75
76
77
78
	src = &get_cpu_var(hv_clock);
	ret = pvclock_clocksource_read(src);
	put_cpu_var(hv_clock);
	return ret;
79
}
80

81
82
83
84
85
86
87
88
89
90
static struct clocksource kvm_clock = {
	.name = "kvm-clock",
	.read = kvm_clock_read,
	.rating = 400,
	.mask = CLOCKSOURCE_MASK(64),
	.mult = 1 << KVM_SCALE,
	.shift = KVM_SCALE,
	.flags = CLOCK_SOURCE_IS_CONTINUOUS,
};

91
static int kvm_register_clock(char *txt)
92
93
94
95
96
{
	int cpu = smp_processor_id();
	int low, high;
	low = (int)__pa(&per_cpu(hv_clock, cpu)) | 1;
	high = ((u64)__pa(&per_cpu(hv_clock, cpu)) >> 32);
97
98
	printk(KERN_INFO "kvm-clock: cpu %d, msr %x:%x, %s\n",
	       cpu, high, low, txt);
99
100
101
	return native_write_msr_safe(MSR_KVM_SYSTEM_TIME, low, high);
}

102
#ifdef CONFIG_X86_LOCAL_APIC
103
104
105
106
107
108
static void kvm_setup_secondary_clock(void)
{
	/*
	 * Now that the first cpu already had this clocksource initialized,
	 * we shouldn't fail.
	 */
109
	WARN_ON(kvm_register_clock("secondary cpu clock"));
110
111
112
	/* ok, done with our trickery, call native */
	setup_secondary_APIC_clock();
}
113
#endif
114

115
116
117
118
119
120
121
122
#ifdef CONFIG_SMP
void __init kvm_smp_prepare_boot_cpu(void)
{
	WARN_ON(kvm_register_clock("primary cpu clock"));
	native_smp_prepare_boot_cpu();
}
#endif

123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
/*
 * After the clock is registered, the host will keep writing to the
 * registered memory location. If the guest happens to shutdown, this memory
 * won't be valid. In cases like kexec, in which you install a new kernel, this
 * means a random memory location will be kept being written. So before any
 * kind of shutdown from our side, we unregister the clock by writting anything
 * that does not have the 'enable' bit set in the msr
 */
#ifdef CONFIG_KEXEC
static void kvm_crash_shutdown(struct pt_regs *regs)
{
	native_write_msr_safe(MSR_KVM_SYSTEM_TIME, 0, 0);
	native_machine_crash_shutdown(regs);
}
#endif

static void kvm_shutdown(void)
{
	native_write_msr_safe(MSR_KVM_SYSTEM_TIME, 0, 0);
	native_machine_shutdown();
}

145
146
147
148
149
150
void __init kvmclock_init(void)
{
	if (!kvm_para_available())
		return;

	if (kvmclock && kvm_para_has_feature(KVM_FEATURE_CLOCKSOURCE)) {
151
		if (kvm_register_clock("boot clock"))
152
153
154
155
			return;
		pv_time_ops.get_wallclock = kvm_get_wallclock;
		pv_time_ops.set_wallclock = kvm_set_wallclock;
		pv_time_ops.sched_clock = kvm_clock_read;
156
#ifdef CONFIG_X86_LOCAL_APIC
157
		pv_apic_ops.setup_secondary_clock = kvm_setup_secondary_clock;
158
159
160
#endif
#ifdef CONFIG_SMP
		smp_ops.smp_prepare_boot_cpu = kvm_smp_prepare_boot_cpu;
161
#endif
162
163
164
165
		machine_ops.shutdown  = kvm_shutdown;
#ifdef CONFIG_KEXEC
		machine_ops.crash_shutdown  = kvm_crash_shutdown;
#endif
166
167
168
		clocksource_register(&kvm_clock);
	}
}