[deliverable/linux.git] / arch / x86 / kernel / vsyscall_64.c

/*
 *  Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE
 *  Copyright 2003 Andi Kleen, SuSE Labs.
 *
 *  Thanks to hpa@transmeta.com for some useful hint.
 *  Special thanks to Ingo Molnar for his early experience with
 *  a different vsyscall implementation for Linux/IA32 and for the name.
 *
 *  vsyscall 1 is located at -10Mbyte, vsyscall 2 is located
 *  at virtual address -10Mbyte+1024bytes etc... There are at max 4
 *  vsyscalls. One vsyscall can reserve more than 1 slot to avoid
 *  jumping out of line if necessary. We cannot add more with this
 *  mechanism because older kernels won't return -ENOSYS.
 *  If we want more than four we need a vDSO.
 *
 *  Note: the concept clashes with user mode linux. If you use UML and
 *  want per guest time just set the kernel.vsyscall64 sysctl to 0.
 */

/* Disable profiling for userspace code: */
#define DISABLE_BRANCH_PROFILING

#include <linux/time.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/timer.h>
#include <linux/seqlock.h>
#include <linux/jiffies.h>
#include <linux/sysctl.h>
#include <linux/clocksource.h>
#include <linux/getcpu.h>
#include <linux/cpu.h>
#include <linux/smp.h>
#include <linux/notifier.h>

#include <asm/vsyscall.h>
#include <asm/pgtable.h>
#include <asm/page.h>
#include <asm/unistd.h>
#include <asm/fixmap.h>
#include <asm/errno.h>
#include <asm/io.h>
#include <asm/segment.h>
#include <asm/desc.h>
#include <asm/topology.h>
#include <asm/vgtod.h>

#define __vsyscall(nr) \
		__attribute__ ((unused, __section__(".vsyscall_" #nr))) notrace
#define __syscall_clobber "r11","cx","memory"

/*
 * vsyscall_gtod_data contains data that is :
 * - readonly from vsyscalls
 * - written by timer interrupt or systcl (/proc/sys/kernel/vsyscall64)
 * Try to keep this structure as small as possible to avoid cache line ping pongs
 */
int __vgetcpu_mode __section_vgetcpu_mode;

struct vsyscall_gtod_data __vsyscall_gtod_data __section_vsyscall_gtod_data =
{
	.lock = SEQLOCK_UNLOCKED,
	.sysctl_enabled = 1,
};

void update_vsyscall_tz(void)
{
	unsigned long flags;

	write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
	/* sys_tz has changed */
	vsyscall_gtod_data.sys_tz = sys_tz;
	write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
}

void update_vsyscall(struct timespec *wall_time, struct clocksource *clock,
		     u32 mult)
{
	unsigned long flags;

	write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
	/* copy vsyscall data */
	vsyscall_gtod_data.clock.vread = clock->vread;
	vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
	vsyscall_gtod_data.clock.mask = clock->mask;
	vsyscall_gtod_data.clock.mult = mult;
	vsyscall_gtod_data.clock.shift = clock->shift;
	vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec;
	vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
	vsyscall_gtod_data.wall_to_monotonic = wall_to_monotonic;
	vsyscall_gtod_data.wall_time_coarse = __current_kernel_time();
	write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
}

/* RED-PEN may want to readd seq locking, but then the variable should be
 * write-once.
 */
static __always_inline void do_get_tz(struct timezone * tz)
{
	*tz = __vsyscall_gtod_data.sys_tz;
}

static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
{
	int ret;
	asm volatile("syscall"
		: "=a" (ret)
		: "0" (__NR_gettimeofday),"D" (tv),"S" (tz)
		: __syscall_clobber );
	return ret;
}

static __always_inline long time_syscall(long *t)
{
	long secs;
	asm volatile("syscall"
		: "=a" (secs)
		: "0" (__NR_time),"D" (t) : __syscall_clobber);
	return secs;
}

static __always_inline void do_vgettimeofday(struct timeval * tv)
{
	cycle_t now, base, mask, cycle_delta;
	unsigned seq;
	unsigned long mult, shift, nsec;
	cycle_t (*vread)(void);
	do {
		seq = read_seqbegin(&__vsyscall_gtod_data.lock);

		vread = __vsyscall_gtod_data.clock.vread;
		if (unlikely(!__vsyscall_gtod_data.sysctl_enabled || !vread)) {
			gettimeofday(tv,NULL);
			return;
		}

		now = vread();
		base = __vsyscall_gtod_data.clock.cycle_last;
		mask = __vsyscall_gtod_data.clock.mask;
		mult = __vsyscall_gtod_data.clock.mult;
		shift = __vsyscall_gtod_data.clock.shift;

		tv->tv_sec = __vsyscall_gtod_data.wall_time_sec;
		nsec = __vsyscall_gtod_data.wall_time_nsec;
	} while (read_seqretry(&__vsyscall_gtod_data.lock, seq));

	/* calculate interval: */
	cycle_delta = (now - base) & mask;
	/* convert to nsecs: */
	nsec += (cycle_delta * mult) >> shift;

	while (nsec >= NSEC_PER_SEC) {
		tv->tv_sec += 1;
		nsec -= NSEC_PER_SEC;
	}
	tv->tv_usec = nsec / NSEC_PER_USEC;
}

int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
{
	if (tv)
		do_vgettimeofday(tv);
	if (tz)
		do_get_tz(tz);
	return 0;
}

/* This will break when the xtime seconds get inaccurate, but that is
 * unlikely */
time_t __vsyscall(1) vtime(time_t *t)
{
	struct timeval tv;
	time_t result;
	if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
		return time_syscall(t);

	vgettimeofday(&tv, NULL);
	result = tv.tv_sec;
	if (t)
		*t = result;
	return result;
}

/* Fast way to get current CPU and node.
   This helps to do per node and per CPU caches in user space.
   The result is not guaranteed without CPU affinity, but usually
   works out because the scheduler tries to keep a thread on the same
   CPU.

   tcache must point to a two element sized long array.
   All arguments can be NULL. */
long __vsyscall(2)
vgetcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache)
{
	unsigned int p;
	unsigned long j = 0;

	/* Fast cache - only recompute value once per jiffies and avoid
	   relatively costly rdtscp/cpuid otherwise.
	   This works because the scheduler usually keeps the process
	   on the same CPU and this syscall doesn't guarantee its
	   results anyways.
	   We do this here because otherwise user space would do it on
	   its own in a likely inferior way (no access to jiffies).
	   If you don't like it pass NULL. */
	if (tcache && tcache->blob[0] == (j = __jiffies)) {
		p = tcache->blob[1];
	} else if (__vgetcpu_mode == VGETCPU_RDTSCP) {
		/* Load per CPU data from RDTSCP */
		native_read_tscp(&p);
	} else {
		/* Load per CPU data from GDT */
		asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
	}
	if (tcache) {
		tcache->blob[0] = j;
		tcache->blob[1] = p;
	}
	if (cpu)
		*cpu = p & 0xfff;
	if (node)
		*node = p >> 12;
	return 0;
}

static long __vsyscall(3) venosys_1(void)
{
	return -ENOSYS;
}

#ifdef CONFIG_SYSCTL
static ctl_table kernel_table2[] = {
	{ .procname = "vsyscall64",
	  .data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
	  .mode = 0644,
	  .proc_handler = proc_dointvec },
	{}
};

static ctl_table kernel_root_table2[] = {
	{ .procname = "kernel", .mode = 0555,
	  .child = kernel_table2 },
	{}
};
#endif

/* Assume __initcall executes before all user space. Hopefully kmod
   doesn't violate that. We'll find out if it does. */
static void __cpuinit vsyscall_set_cpu(int cpu)
{
	unsigned long d;
	unsigned long node = 0;
#ifdef CONFIG_NUMA
	node = cpu_to_node(cpu);
#endif
	if (cpu_has(&cpu_data(cpu), X86_FEATURE_RDTSCP))
		write_rdtscp_aux((node << 12) | cpu);

	/* Store cpu number in limit so that it can be loaded quickly
	   in user space in vgetcpu.
	   12 bits for the CPU and 8 bits for the node. */
	d = 0x0f40000000000ULL;
	d |= cpu;
	d |= (node & 0xf) << 12;
	d |= (node >> 4) << 48;
	write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_PER_CPU, &d, DESCTYPE_S);
}

static void __cpuinit cpu_vsyscall_init(void *arg)
{
	/* preemption should be already off */
	vsyscall_set_cpu(raw_smp_processor_id());
}

static int __cpuinit
cpu_vsyscall_notifier(struct notifier_block *n, unsigned long action, void *arg)
{
	long cpu = (long)arg;
	if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN)
		smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 1);
	return NOTIFY_DONE;
}

void __init map_vsyscall(void)
{
	extern char __vsyscall_0;
	unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);

	/* Note that VSYSCALL_MAPPED_PAGES must agree with the code below. */
	__set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
}

static int __init vsyscall_init(void)
{
	BUG_ON(((unsigned long) &vgettimeofday !=
			VSYSCALL_ADDR(__NR_vgettimeofday)));
	BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
	BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
	BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
#ifdef CONFIG_SYSCTL
	register_sysctl_table(kernel_root_table2);
#endif
	on_each_cpu(cpu_vsyscall_init, NULL, 1);
	/* notifier priority > KVM */
	hotcpu_notifier(cpu_vsyscall_notifier, 30);
	return 0;
}

__initcall(vsyscall_init);
Commit	Line	Data
1da177e4	1	/*
1da177e4 LT	2	* Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE
	3	* Copyright 2003 Andi Kleen, SuSE Labs.
	4	*
	5	* Thanks to hpa@transmeta.com for some useful hint.
	6	* Special thanks to Ingo Molnar for his early experience with
	7	* a different vsyscall implementation for Linux/IA32 and for the name.
	8	*
	9	* vsyscall 1 is located at -10Mbyte, vsyscall 2 is located
	10	* at virtual address -10Mbyte+1024bytes etc... There are at max 4
	11	* vsyscalls. One vsyscall can reserve more than 1 slot to avoid
	12	* jumping out of line if necessary. We cannot add more with this
	13	* mechanism because older kernels won't return -ENOSYS.
	14	* If we want more than four we need a vDSO.
	15	*
	16	* Note: the concept clashes with user mode linux. If you use UML and
	17	* want per guest time just set the kernel.vsyscall64 sysctl to 0.
	18	*/
	19
2b7d0390	20	/* Disable profiling for userspace code: */
2ed84eeb	21	#define DISABLE_BRANCH_PROFILING
1f0d69a9	22
1da177e4 LT	23	#include <linux/time.h>
	24	#include <linux/init.h>
	25	#include <linux/kernel.h>
	26	#include <linux/timer.h>
	27	#include <linux/seqlock.h>
	28	#include <linux/jiffies.h>
	29	#include <linux/sysctl.h>
7460ed28	30	#include <linux/clocksource.h>
c08c8205	31	#include <linux/getcpu.h>
8c131af1 AK	32	#include <linux/cpu.h>
	33	#include <linux/smp.h>
	34	#include <linux/notifier.h>
1da177e4 LT	35
	36	#include <asm/vsyscall.h>
	37	#include <asm/pgtable.h>
	38	#include <asm/page.h>
7460ed28	39	#include <asm/unistd.h>
1da177e4 LT	40	#include <asm/fixmap.h>
	41	#include <asm/errno.h>
	42	#include <asm/io.h>
c08c8205 VP	43	#include <asm/segment.h>
	44	#include <asm/desc.h>
	45	#include <asm/topology.h>
2aae950b	46	#include <asm/vgtod.h>
1da177e4	47
23adec55 SR	48	#define __vsyscall(nr) \
23adec55 SR	49	__attribute__ ((unused, __section__(".vsyscall_" #nr))) notrace
65ea5b03	50	#define __syscall_clobber "r11","cx","memory"
1da177e4	51
c8118c6c ED	52	/*
	53	* vsyscall_gtod_data contains data that is :
	54	* - readonly from vsyscalls
676b1855	55	* - written by timer interrupt or systcl (/proc/sys/kernel/vsyscall64)
c8118c6c ED	56	* Try to keep this structure as small as possible to avoid cache line ping pongs
c8118c6c ED	57	*/
c08c8205	58	int __vgetcpu_mode __section_vgetcpu_mode;
1da177e4	59
2aae950b	60	struct vsyscall_gtod_data __vsyscall_gtod_data __section_vsyscall_gtod_data =
1da177e4	61	{
7460ed28	62	.lock = SEQLOCK_UNLOCKED,
	63	.sysctl_enabled = 1,
	64	};
1da177e4	65
2c622148 TB	66	void update_vsyscall_tz(void)
	67	{
	68	unsigned long flags;
	69
	70	write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
	71	/* sys_tz has changed */
	72	vsyscall_gtod_data.sys_tz = sys_tz;
	73	write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
	74	}
	75
0696b711 LM	76	void update_vsyscall(struct timespec wall_time, struct clocksource clock,
0696b711 LM	77	u32 mult)
1da177e4	78	{
7460ed28	79	unsigned long flags;
1da177e4	80
7460ed28	81	write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
7460ed28	82	/* copy vsyscall data */
c8118c6c ED	83	vsyscall_gtod_data.clock.vread = clock->vread;
	84	vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
	85	vsyscall_gtod_data.clock.mask = clock->mask;
0696b711	86	vsyscall_gtod_data.clock.mult = mult;
c8118c6c ED	87	vsyscall_gtod_data.clock.shift = clock->shift;
	88	vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec;
	89	vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
2aae950b	90	vsyscall_gtod_data.wall_to_monotonic = wall_to_monotonic;
da15cfda	91	vsyscall_gtod_data.wall_time_coarse = __current_kernel_time();
7460ed28	92	write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
1da177e4 LT	93	}
1da177e4 LT	94
7460ed28	95	/* RED-PEN may want to readd seq locking, but then the variable should be
	96	* write-once.
	97	*/
2c8bc944	98	static __always_inline void do_get_tz(struct timezone * tz)
1da177e4	99	{
7460ed28	100	*tz = __vsyscall_gtod_data.sys_tz;
1da177e4 LT	101	}
1da177e4 LT	102
2c8bc944	103	static __always_inline int gettimeofday(struct timeval tv, struct timezone tz)
1da177e4 LT	104	{
1da177e4 LT	105	int ret;
ce28b986	106	asm volatile("syscall"
1da177e4	107	: "=a" (ret)
7460ed28	108	: "0" (__NR_gettimeofday),"D" (tv),"S" (tz)
7460ed28	109	: __syscall_clobber );
1da177e4 LT	110	return ret;
	111	}
	112
2c8bc944	113	static __always_inline long time_syscall(long *t)
1da177e4 LT	114	{
1da177e4 LT	115	long secs;
ce28b986	116	asm volatile("syscall"
1da177e4 LT	117	: "=a" (secs)
	118	: "0" (__NR_time),"D" (t) : __syscall_clobber);
	119	return secs;
	120	}
	121
7460ed28	122	static __always_inline void do_vgettimeofday(struct timeval * tv)
	123	{
	124	cycle_t now, base, mask, cycle_delta;
c8118c6c ED	125	unsigned seq;
c8118c6c ED	126	unsigned long mult, shift, nsec;
7460ed28	127	cycle_t (*vread)(void);
	128	do {
	129	seq = read_seqbegin(&__vsyscall_gtod_data.lock);
	130
	131	vread = __vsyscall_gtod_data.clock.vread;
	132	if (unlikely(!__vsyscall_gtod_data.sysctl_enabled \|\| !vread)) {
89952d13	133	gettimeofday(tv,NULL);
7460ed28	134	return;
7460ed28	135	}
cb9e35dc	136
7460ed28	137	now = vread();
	138	base = __vsyscall_gtod_data.clock.cycle_last;
	139	mask = __vsyscall_gtod_data.clock.mask;
	140	mult = __vsyscall_gtod_data.clock.mult;
	141	shift = __vsyscall_gtod_data.clock.shift;
	142
c8118c6c ED	143	tv->tv_sec = __vsyscall_gtod_data.wall_time_sec;
c8118c6c ED	144	nsec = __vsyscall_gtod_data.wall_time_nsec;
7460ed28	145	} while (read_seqretry(&__vsyscall_gtod_data.lock, seq));
	146
	147	/* calculate interval: */
	148	cycle_delta = (now - base) & mask;
	149	/* convert to nsecs: */
c8118c6c	150	nsec += (cycle_delta * mult) >> shift;
7460ed28	151
c8118c6c	152	while (nsec >= NSEC_PER_SEC) {
7460ed28	153	tv->tv_sec += 1;
c8118c6c	154	nsec -= NSEC_PER_SEC;
7460ed28	155	}
c8118c6c	156	tv->tv_usec = nsec / NSEC_PER_USEC;
7460ed28	157	}
7460ed28	158
2e8ad43e	159	int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
1da177e4	160	{
1da177e4 LT	161	if (tv)
	162	do_vgettimeofday(tv);
	163	if (tz)
	164	do_get_tz(tz);
	165	return 0;
	166	}
	167
	168	/* This will break when the xtime seconds get inaccurate, but that is
	169	* unlikely */
2e8ad43e	170	time_t __vsyscall(1) vtime(time_t *t)
1da177e4	171	{
d0aff6e6	172	struct timeval tv;
272a3713	173	time_t result;
7460ed28	174	if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
1da177e4	175	return time_syscall(t);
d0aff6e6	176
c80544dc	177	vgettimeofday(&tv, NULL);
d0aff6e6	178	result = tv.tv_sec;
272a3713 ED	179	if (t)
	180	*t = result;
	181	return result;
1da177e4 LT	182	}
1da177e4 LT	183
c08c8205 VP	184	/* Fast way to get current CPU and node.
	185	This helps to do per node and per CPU caches in user space.
	186	The result is not guaranteed without CPU affinity, but usually
	187	works out because the scheduler tries to keep a thread on the same
	188	CPU.
	189
	190	tcache must point to a two element sized long array.
	191	All arguments can be NULL. */
	192	long __vsyscall(2)
	193	vgetcpu(unsigned cpu, unsigned node, struct getcpu_cache *tcache)
1da177e4	194	{
8f12dea6	195	unsigned int p;
c08c8205 VP	196	unsigned long j = 0;
	197
	198	/* Fast cache - only recompute value once per jiffies and avoid
	199	relatively costly rdtscp/cpuid otherwise.
	200	This works because the scheduler usually keeps the process
	201	on the same CPU and this syscall doesn't guarantee its
	202	results anyways.
	203	We do this here because otherwise user space would do it on
	204	its own in a likely inferior way (no access to jiffies).
	205	If you don't like it pass NULL. */
34596dc9 AK	206	if (tcache && tcache->blob[0] == (j = __jiffies)) {
34596dc9 AK	207	p = tcache->blob[1];
c08c8205 VP	208	} else if (__vgetcpu_mode == VGETCPU_RDTSCP) {
c08c8205 VP	209	/* Load per CPU data from RDTSCP */
8f12dea6	210	native_read_tscp(&p);
c08c8205 VP	211	} else {
	212	/* Load per CPU data from GDT */
	213	asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
	214	}
	215	if (tcache) {
34596dc9 AK	216	tcache->blob[0] = j;
34596dc9 AK	217	tcache->blob[1] = p;
c08c8205 VP	218	}
	219	if (cpu)
	220	*cpu = p & 0xfff;
	221	if (node)
	222	*node = p >> 12;
	223	return 0;
1da177e4 LT	224	}
1da177e4 LT	225
a4928cff	226	static long __vsyscall(3) venosys_1(void)
1da177e4 LT	227	{
	228	return -ENOSYS;
	229	}
	230
	231	#ifdef CONFIG_SYSCTL
1da177e4	232	static ctl_table kernel_table2[] = {
282a821f	233	{ .procname = "vsyscall64",
7460ed28	234	.data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
d67bbacb	235	.mode = 0644,
8d65af78	236	.proc_handler = proc_dointvec },
7a44d37d	237	{}
1da177e4 LT	238	};
	239
	240	static ctl_table kernel_root_table2[] = {
24a06562	241	{ .procname = "kernel", .mode = 0555,
1da177e4	242	.child = kernel_table2 },
7a44d37d	243	{}
1da177e4	244	};
1da177e4 LT	245	#endif
1da177e4 LT	246
8c131af1 AK	247	/* Assume __initcall executes before all user space. Hopefully kmod
	248	doesn't violate that. We'll find out if it does. */
	249	static void __cpuinit vsyscall_set_cpu(int cpu)
c08c8205	250	{
fc8b8a60	251	unsigned long d;
c08c8205 VP	252	unsigned long node = 0;
c08c8205 VP	253	#ifdef CONFIG_NUMA
98c9e27a	254	node = cpu_to_node(cpu);
c08c8205	255	#endif
92cb7612	256	if (cpu_has(&cpu_data(cpu), X86_FEATURE_RDTSCP))
8c131af1	257	write_rdtscp_aux((node << 12) \| cpu);
c08c8205 VP	258
	259	/* Store cpu number in limit so that it can be loaded quickly
	260	in user space in vgetcpu.
	261	12 bits for the CPU and 8 bits for the node. */
fc8b8a60 JF	262	d = 0x0f40000000000ULL;
	263	d \|= cpu;
	264	d \|= (node & 0xf) << 12;
	265	d \|= (node >> 4) << 48;
	266	write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_PER_CPU, &d, DESCTYPE_S);
c08c8205 VP	267	}
c08c8205 VP	268
8c131af1 AK	269	static void __cpuinit cpu_vsyscall_init(void *arg)
	270	{
	271	/* preemption should be already off */
	272	vsyscall_set_cpu(raw_smp_processor_id());
	273	}
	274
	275	static int __cpuinit
	276	cpu_vsyscall_notifier(struct notifier_block n, unsigned long action, void arg)
	277	{
	278	long cpu = (long)arg;
8bb78442	279	if (action == CPU_ONLINE \|\| action == CPU_ONLINE_FROZEN)
8691e5a8	280	smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 1);
8c131af1 AK	281	return NOTIFY_DONE;
	282	}
	283
e4026440	284	void __init map_vsyscall(void)
1da177e4 LT	285	{
	286	extern char __vsyscall_0;
	287	unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
	288
103efcd9	289	/* Note that VSYSCALL_MAPPED_PAGES must agree with the code below. */
1da177e4 LT	290	__set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
	291	}
	292
	293	static int __init vsyscall_init(void)
	294	{
	295	BUG_ON(((unsigned long) &vgettimeofday !=
	296	VSYSCALL_ADDR(__NR_vgettimeofday)));
	297	BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
	298	BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
c08c8205	299	BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
f3c5f5e7	300	#ifdef CONFIG_SYSCTL
0b4d4147	301	register_sysctl_table(kernel_root_table2);
f3c5f5e7	302	#endif
15c8b6c1	303	on_each_cpu(cpu_vsyscall_init, NULL, 1);
be43f83d SY	304	/* notifier priority > KVM */
be43f83d SY	305	hotcpu_notifier(cpu_vsyscall_notifier, 30);
1da177e4 LT	306	return 0;
	307	}
	308
	309	__initcall(vsyscall_init);