[deliverable/linux.git] / arch / arm / xen / enlighten.c

#include <xen/xen.h>
#include <xen/events.h>
#include <xen/grant_table.h>
#include <xen/hvm.h>
#include <xen/interface/vcpu.h>
#include <xen/interface/xen.h>
#include <xen/interface/memory.h>
#include <xen/interface/hvm/params.h>
#include <xen/features.h>
#include <xen/platform_pci.h>
#include <xen/xenbus.h>
#include <xen/page.h>
#include <xen/interface/sched.h>
#include <xen/xen-ops.h>
#include <asm/paravirt.h>
#include <asm/xen/hypervisor.h>
#include <asm/xen/hypercall.h>
#include <asm/system_misc.h>
#include <linux/interrupt.h>
#include <linux/irqreturn.h>
#include <linux/module.h>
#include <linux/of.h>
#include <linux/of_irq.h>
#include <linux/of_address.h>
#include <linux/cpuidle.h>
#include <linux/cpufreq.h>
#include <linux/cpu.h>
#include <linux/console.h>
#include <linux/pvclock_gtod.h>
#include <linux/time64.h>
#include <linux/timekeeping.h>
#include <linux/timekeeper_internal.h>
#include <linux/acpi.h>

#include <linux/mm.h>

struct start_info _xen_start_info;
struct start_info *xen_start_info = &_xen_start_info;
EXPORT_SYMBOL(xen_start_info);

enum xen_domain_type xen_domain_type = XEN_NATIVE;
EXPORT_SYMBOL(xen_domain_type);

struct shared_info xen_dummy_shared_info;
struct shared_info *HYPERVISOR_shared_info = (void *)&xen_dummy_shared_info;

DEFINE_PER_CPU(struct vcpu_info *, xen_vcpu);
static struct vcpu_info __percpu *xen_vcpu_info;

/* These are unused until we support booting "pre-ballooned" */
unsigned long xen_released_pages;
struct xen_memory_region xen_extra_mem[XEN_EXTRA_MEM_MAX_REGIONS] __initdata;

static __read_mostly unsigned int xen_events_irq;

static __initdata struct device_node *xen_node;

int xen_remap_domain_gfn_array(struct vm_area_struct *vma,
			       unsigned long addr,
			       xen_pfn_t *gfn, int nr,
			       int *err_ptr, pgprot_t prot,
			       unsigned domid,
			       struct page **pages)
{
	return xen_xlate_remap_gfn_array(vma, addr, gfn, nr, err_ptr,
					 prot, domid, pages);
}
EXPORT_SYMBOL_GPL(xen_remap_domain_gfn_array);

/* Not used by XENFEAT_auto_translated guests. */
int xen_remap_domain_gfn_range(struct vm_area_struct *vma,
                              unsigned long addr,
                              xen_pfn_t gfn, int nr,
                              pgprot_t prot, unsigned domid,
                              struct page **pages)
{
	return -ENOSYS;
}
EXPORT_SYMBOL_GPL(xen_remap_domain_gfn_range);

int xen_unmap_domain_gfn_range(struct vm_area_struct *vma,
			       int nr, struct page **pages)
{
	return xen_xlate_unmap_gfn_range(vma, nr, pages);
}
EXPORT_SYMBOL_GPL(xen_unmap_domain_gfn_range);

static unsigned long long xen_stolen_accounting(int cpu)
{
	struct vcpu_runstate_info state;

	BUG_ON(cpu != smp_processor_id());

	xen_get_runstate_snapshot(&state);

	WARN_ON(state.state != RUNSTATE_running);

	return state.time[RUNSTATE_runnable] + state.time[RUNSTATE_offline];
}

static void xen_read_wallclock(struct timespec64 *ts)
{
	u32 version;
	struct timespec64 now, ts_monotonic;
	struct shared_info *s = HYPERVISOR_shared_info;
	struct pvclock_wall_clock *wall_clock = &(s->wc);

	/* get wallclock at system boot */
	do {
		version = wall_clock->version;
		rmb();		/* fetch version before time */
		now.tv_sec  = ((uint64_t)wall_clock->sec_hi << 32) | wall_clock->sec;
		now.tv_nsec = wall_clock->nsec;
		rmb();		/* fetch time before checking version */
	} while ((wall_clock->version & 1) || (version != wall_clock->version));

	/* time since system boot */
	ktime_get_ts64(&ts_monotonic);
	*ts = timespec64_add(now, ts_monotonic);
}

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, system_time;
	struct timekeeper *tk = priv;

	now.tv_sec = tk->xtime_sec;
	now.tv_nsec = (long)(tk->tkr_mono.xtime_nsec >> tk->tkr_mono.shift);
	system_time = timespec64_add(now, tk->wall_to_monotonic);

	/*
	 * 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;

	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 = timespec64_to_ns(&system_time);
	(void)HYPERVISOR_platform_op(&op);

	/*
	 * 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 void xen_percpu_init(void)
{
	struct vcpu_register_vcpu_info info;
	struct vcpu_info *vcpup;
	int err;
	int cpu = get_cpu();

	/* 
	 * VCPUOP_register_vcpu_info cannot be called twice for the same
	 * vcpu, so if vcpu_info is already registered, just get out. This
	 * can happen with cpu-hotplug.
	 */
	if (per_cpu(xen_vcpu, cpu) != NULL)
		goto after_register_vcpu_info;

	pr_info("Xen: initializing cpu%d\n", cpu);
	vcpup = per_cpu_ptr(xen_vcpu_info, cpu);

	info.mfn = virt_to_gfn(vcpup);
	info.offset = xen_offset_in_page(vcpup);

	err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, cpu, &info);
	BUG_ON(err);
	per_cpu(xen_vcpu, cpu) = vcpup;

	xen_setup_runstate_info(cpu);

after_register_vcpu_info:
	enable_percpu_irq(xen_events_irq, 0);
	put_cpu();
}

static void xen_restart(enum reboot_mode reboot_mode, const char *cmd)
{
	struct sched_shutdown r = { .reason = SHUTDOWN_reboot };
	int rc;
	rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &r);
	BUG_ON(rc);
}

static void xen_power_off(void)
{
	struct sched_shutdown r = { .reason = SHUTDOWN_poweroff };
	int rc;
	rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &r);
	BUG_ON(rc);
}

static int xen_cpu_notification(struct notifier_block *self,
				unsigned long action,
				void *hcpu)
{
	switch (action) {
	case CPU_STARTING:
		xen_percpu_init();
		break;
	case CPU_DYING:
		disable_percpu_irq(xen_events_irq);
		break;
	default:
		break;
	}

	return NOTIFY_OK;
}

static struct notifier_block xen_cpu_notifier = {
	.notifier_call = xen_cpu_notification,
};

static irqreturn_t xen_arm_callback(int irq, void *arg)
{
	xen_hvm_evtchn_do_upcall();
	return IRQ_HANDLED;
}

/*
 * see Documentation/devicetree/bindings/arm/xen.txt for the
 * documentation of the Xen Device Tree format.
 */
#define GRANT_TABLE_PHYSADDR 0
void __init xen_early_init(void)
{
	int len;
	const char *s = NULL;
	const char *version = NULL;
	const char *xen_prefix = "xen,xen-";

	xen_node = of_find_compatible_node(NULL, NULL, "xen,xen");
	if (!xen_node) {
		pr_debug("No Xen support\n");
		return;
	}
	s = of_get_property(xen_node, "compatible", &len);
	if (strlen(xen_prefix) + 3  < len &&
			!strncmp(xen_prefix, s, strlen(xen_prefix)))
		version = s + strlen(xen_prefix);
	if (version == NULL) {
		pr_debug("Xen version not found\n");
		return;
	}

	pr_info("Xen %s support found\n", version);

	xen_domain_type = XEN_HVM_DOMAIN;

	xen_setup_features();

	if (xen_feature(XENFEAT_dom0))
		xen_start_info->flags |= SIF_INITDOMAIN|SIF_PRIVILEGED;
	else
		xen_start_info->flags &= ~(SIF_INITDOMAIN|SIF_PRIVILEGED);

	if (!console_set_on_cmdline && !xen_initial_domain())
		add_preferred_console("hvc", 0, NULL);
}

static void __init xen_acpi_guest_init(void)
{
#ifdef CONFIG_ACPI
	struct xen_hvm_param a;
	int interrupt, trigger, polarity;

	a.domid = DOMID_SELF;
	a.index = HVM_PARAM_CALLBACK_IRQ;

	if (HYPERVISOR_hvm_op(HVMOP_get_param, &a)
	    || (a.value >> 56) != HVM_PARAM_CALLBACK_TYPE_PPI) {
		xen_events_irq = 0;
		return;
	}

	interrupt = a.value & 0xff;
	trigger = ((a.value >> 8) & 0x1) ? ACPI_EDGE_SENSITIVE
					 : ACPI_LEVEL_SENSITIVE;
	polarity = ((a.value >> 8) & 0x2) ? ACPI_ACTIVE_LOW
					  : ACPI_ACTIVE_HIGH;
	xen_events_irq = acpi_register_gsi(NULL, interrupt, trigger, polarity);
#endif
}

static void __init xen_dt_guest_init(void)
{
	xen_events_irq = irq_of_parse_and_map(xen_node, 0);
}

static int __init xen_guest_init(void)
{
	struct xen_add_to_physmap xatp;
	struct shared_info *shared_info_page = NULL;

	if (!xen_domain())
		return 0;

	if (!acpi_disabled)
		xen_acpi_guest_init();
	else
		xen_dt_guest_init();

	if (!xen_events_irq) {
		pr_err("Xen event channel interrupt not found\n");
		return -ENODEV;
	}

	shared_info_page = (struct shared_info *)get_zeroed_page(GFP_KERNEL);

	if (!shared_info_page) {
		pr_err("not enough memory\n");
		return -ENOMEM;
	}
	xatp.domid = DOMID_SELF;
	xatp.idx = 0;
	xatp.space = XENMAPSPACE_shared_info;
	xatp.gpfn = virt_to_gfn(shared_info_page);
	if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
		BUG();

	HYPERVISOR_shared_info = (struct shared_info *)shared_info_page;

	/* xen_vcpu is a pointer to the vcpu_info struct in the shared_info
	 * page, we use it in the event channel upcall and in some pvclock
	 * related functions. 
	 * The shared info contains exactly 1 CPU (the boot CPU). The guest
	 * is required to use VCPUOP_register_vcpu_info to place vcpu info
	 * for secondary CPUs as they are brought up.
	 * For uniformity we use VCPUOP_register_vcpu_info even on cpu0.
	 */
	xen_vcpu_info = __alloc_percpu(sizeof(struct vcpu_info),
			                       sizeof(struct vcpu_info));
	if (xen_vcpu_info == NULL)
		return -ENOMEM;

	xen_auto_xlat_grant_frames.count = gnttab_max_grant_frames();
	if (xen_xlate_map_ballooned_pages(&xen_auto_xlat_grant_frames.pfn,
					  &xen_auto_xlat_grant_frames.vaddr,
					  xen_auto_xlat_grant_frames.count)) {
		free_percpu(xen_vcpu_info);
		return -ENOMEM;
	}
	gnttab_init();
	if (!xen_initial_domain())
		xenbus_probe(NULL);

	/*
	 * Making sure board specific code will not set up ops for
	 * cpu idle and cpu freq.
	 */
	disable_cpuidle();
	disable_cpufreq();

	xen_init_IRQ();

	if (request_percpu_irq(xen_events_irq, xen_arm_callback,
			       "events", &xen_vcpu)) {
		pr_err("Error request IRQ %d\n", xen_events_irq);
		return -EINVAL;
	}

	xen_percpu_init();

	register_cpu_notifier(&xen_cpu_notifier);

	pv_time_ops.steal_clock = xen_stolen_accounting;
	static_key_slow_inc(&paravirt_steal_enabled);
	if (xen_initial_domain())
		pvclock_gtod_register_notifier(&xen_pvclock_gtod_notifier);

	return 0;
}
early_initcall(xen_guest_init);

static int __init xen_pm_init(void)
{
	if (!xen_domain())
		return -ENODEV;

	pm_power_off = xen_power_off;
	arm_pm_restart = xen_restart;
	if (!xen_initial_domain()) {
		struct timespec64 ts;
		xen_read_wallclock(&ts);
		do_settimeofday64(&ts);
	}

	return 0;
}
late_initcall(xen_pm_init);


/* empty stubs */
void xen_arch_pre_suspend(void) { }
void xen_arch_post_suspend(int suspend_cancelled) { }
void xen_timer_resume(void) { }
void xen_arch_resume(void) { }
void xen_arch_suspend(void) { }


/* In the hypercall.S file. */
EXPORT_SYMBOL_GPL(HYPERVISOR_event_channel_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_grant_table_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_xen_version);
EXPORT_SYMBOL_GPL(HYPERVISOR_console_io);
EXPORT_SYMBOL_GPL(HYPERVISOR_sched_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_hvm_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
EXPORT_SYMBOL_GPL(privcmd_call);
Commit	Line	Data
4c071ee5	1	#include <xen/xen.h>
0ec53ecf	2	#include <xen/events.h>
b3b52fd8 SS	3	#include <xen/grant_table.h>
b3b52fd8 SS	4	#include <xen/hvm.h>
9a9ab3cc	5	#include <xen/interface/vcpu.h>
4c071ee5 SS	6	#include <xen/interface/xen.h>
4c071ee5 SS	7	#include <xen/interface/memory.h>
b3b52fd8	8	#include <xen/interface/hvm/params.h>
ef61ee0d	9	#include <xen/features.h>
4c071ee5	10	#include <xen/platform_pci.h>
b3b52fd8	11	#include <xen/xenbus.h>
c61ba729	12	#include <xen/page.h>
6abb749e	13	#include <xen/interface/sched.h>
f832da06	14	#include <xen/xen-ops.h>
34e38523	15	#include <asm/paravirt.h>
4c071ee5 SS	16	#include <asm/xen/hypervisor.h>
4c071ee5 SS	17	#include <asm/xen/hypercall.h>
6abb749e	18	#include <asm/system_misc.h>
0ec53ecf SS	19	#include <linux/interrupt.h>
0ec53ecf SS	20	#include <linux/irqreturn.h>
4c071ee5	21	#include <linux/module.h>
2e01f166 SS	22	#include <linux/of.h>
	23	#include <linux/of_irq.h>
	24	#include <linux/of_address.h>
e1a9c16b JG	25	#include <linux/cpuidle.h>
e1a9c16b JG	26	#include <linux/cpufreq.h>
8b271d57	27	#include <linux/cpu.h>
f1dddd11	28	#include <linux/console.h>
7d5f6f81 SS	29	#include <linux/pvclock_gtod.h>
7d5f6f81 SS	30	#include <linux/time64.h>
e709fba1	31	#include <linux/timekeeping.h>
7d5f6f81	32	#include <linux/timekeeper_internal.h>
d22cbe65	33	#include <linux/acpi.h>
4c071ee5	34
f832da06 IC	35	#include <linux/mm.h>
f832da06 IC	36
4c071ee5 SS	37	struct start_info _xen_start_info;
4c071ee5 SS	38	struct start_info *xen_start_info = &_xen_start_info;
35c8ab4c	39	EXPORT_SYMBOL(xen_start_info);
4c071ee5 SS	40
4c071ee5 SS	41	enum xen_domain_type xen_domain_type = XEN_NATIVE;
35c8ab4c	42	EXPORT_SYMBOL(xen_domain_type);
4c071ee5 SS	43
	44	struct shared_info xen_dummy_shared_info;
	45	struct shared_info HYPERVISOR_shared_info = (void )&xen_dummy_shared_info;
	46
	47	DEFINE_PER_CPU(struct vcpu_info *, xen_vcpu);
9a9ab3cc	48	static struct vcpu_info __percpu *xen_vcpu_info;
4c071ee5	49
c61ba729 IC	50	/* These are unused until we support booting "pre-ballooned" */
	51	unsigned long xen_released_pages;
	52	struct xen_memory_region xen_extra_mem[XEN_EXTRA_MEM_MAX_REGIONS] __initdata;
	53
81e863c3	54	static __read_mostly unsigned int xen_events_irq;
0ec53ecf	55
5882bfef SS	56	static __initdata struct device_node *xen_node;
5882bfef SS	57
a13d7201	58	int xen_remap_domain_gfn_array(struct vm_area_struct *vma,
4c071ee5	59	unsigned long addr,
a13d7201	60	xen_pfn_t *gfn, int nr,
4e8c0c8c DV	61	int *err_ptr, pgprot_t prot,
4e8c0c8c DV	62	unsigned domid,
f832da06	63	struct page **pages)
4c071ee5	64	{
a13d7201	65	return xen_xlate_remap_gfn_array(vma, addr, gfn, nr, err_ptr,
628c28ee	66	prot, domid, pages);
4c071ee5	67	}
a13d7201	68	EXPORT_SYMBOL_GPL(xen_remap_domain_gfn_array);
4e8c0c8c DV	69
4e8c0c8c DV	70	/* Not used by XENFEAT_auto_translated guests. */
a13d7201	71	int xen_remap_domain_gfn_range(struct vm_area_struct *vma,
4e8c0c8c	72	unsigned long addr,
a13d7201	73	xen_pfn_t gfn, int nr,
4e8c0c8c DV	74	pgprot_t prot, unsigned domid,
	75	struct page **pages)
	76	{
	77	return -ENOSYS;
	78	}
a13d7201	79	EXPORT_SYMBOL_GPL(xen_remap_domain_gfn_range);
2e01f166	80
a13d7201	81	int xen_unmap_domain_gfn_range(struct vm_area_struct *vma,
f832da06 IC	82	int nr, struct page **pages)
f832da06 IC	83	{
628c28ee	84	return xen_xlate_unmap_gfn_range(vma, nr, pages);
f832da06	85	}
a13d7201	86	EXPORT_SYMBOL_GPL(xen_unmap_domain_gfn_range);
f832da06	87
34e38523 SS	88	static unsigned long long xen_stolen_accounting(int cpu)
	89	{
	90	struct vcpu_runstate_info state;
	91
	92	BUG_ON(cpu != smp_processor_id());
	93
	94	xen_get_runstate_snapshot(&state);
	95
	96	WARN_ON(state.state != RUNSTATE_running);
	97
	98	return state.time[RUNSTATE_runnable] + state.time[RUNSTATE_offline];
	99	}
	100
e709fba1 SS	101	static void xen_read_wallclock(struct timespec64 *ts)
	102	{
	103	u32 version;
	104	struct timespec64 now, ts_monotonic;
	105	struct shared_info *s = HYPERVISOR_shared_info;
	106	struct pvclock_wall_clock *wall_clock = &(s->wc);
	107
	108	/* get wallclock at system boot */
	109	do {
	110	version = wall_clock->version;
	111	rmb(); /* fetch version before time */
	112	now.tv_sec = ((uint64_t)wall_clock->sec_hi << 32) \| wall_clock->sec;
	113	now.tv_nsec = wall_clock->nsec;
	114	rmb(); /* fetch time before checking version */
	115	} while ((wall_clock->version & 1) \|\| (version != wall_clock->version));
	116
	117	/* time since system boot */
	118	ktime_get_ts64(&ts_monotonic);
	119	*ts = timespec64_add(now, ts_monotonic);
	120	}
	121
7d5f6f81 SS	122	static int xen_pvclock_gtod_notify(struct notifier_block *nb,
	123	unsigned long was_set, void *priv)
	124	{
	125	/* Protected by the calling core code serialization */
	126	static struct timespec64 next_sync;
	127
	128	struct xen_platform_op op;
	129	struct timespec64 now, system_time;
	130	struct timekeeper *tk = priv;
	131
	132	now.tv_sec = tk->xtime_sec;
	133	now.tv_nsec = (long)(tk->tkr_mono.xtime_nsec >> tk->tkr_mono.shift);
	134	system_time = timespec64_add(now, tk->wall_to_monotonic);
	135
	136	/*
	137	* We only take the expensive HV call when the clock was set
	138	* or when the 11 minutes RTC synchronization time elapsed.
	139	*/
	140	if (!was_set && timespec64_compare(&now, &next_sync) < 0)
	141	return NOTIFY_OK;
	142
	143	op.cmd = XENPF_settime64;
	144	op.u.settime64.mbz = 0;
	145	op.u.settime64.secs = now.tv_sec;
	146	op.u.settime64.nsecs = now.tv_nsec;
	147	op.u.settime64.system_time = timespec64_to_ns(&system_time);
	148	(void)HYPERVISOR_platform_op(&op);
	149
	150	/*
	151	* Move the next drift compensation time 11 minutes
	152	* ahead. That's emulating the sync_cmos_clock() update for
	153	* the hardware RTC.
	154	*/
	155	next_sync = now;
	156	next_sync.tv_sec += 11 * 60;
	157
	158	return NOTIFY_OK;
	159	}
	160
	161	static struct notifier_block xen_pvclock_gtod_notifier = {
	162	.notifier_call = xen_pvclock_gtod_notify,
	163	};
	164
8b271d57	165	static void xen_percpu_init(void)
9a9ab3cc SS	166	{
	167	struct vcpu_register_vcpu_info info;
	168	struct vcpu_info *vcpup;
	169	int err;
3cc8e40e	170	int cpu = get_cpu();
9a9ab3cc	171
cb9644bf SS	172	/*
	173	* VCPUOP_register_vcpu_info cannot be called twice for the same
	174	* vcpu, so if vcpu_info is already registered, just get out. This
	175	* can happen with cpu-hotplug.
	176	*/
	177	if (per_cpu(xen_vcpu, cpu) != NULL)
	178	goto after_register_vcpu_info;
	179
9a9ab3cc SS	180	pr_info("Xen: initializing cpu%d\n", cpu);
	181	vcpup = per_cpu_ptr(xen_vcpu_info, cpu);
	182
250c9af3 JG	183	info.mfn = virt_to_gfn(vcpup);
250c9af3 JG	184	info.offset = xen_offset_in_page(vcpup);
9a9ab3cc SS	185
9a9ab3cc SS	186	err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, cpu, &info);
d7266d78 SS	187	BUG_ON(err);
	188	per_cpu(xen_vcpu, cpu) = vcpup;
	189
34e38523 SS	190	xen_setup_runstate_info(cpu);
34e38523 SS	191
cb9644bf	192	after_register_vcpu_info:
3cc8e40e	193	enable_percpu_irq(xen_events_irq, 0);
0d7febe5	194	put_cpu();
9a9ab3cc SS	195	}
9a9ab3cc SS	196
2451ade0	197	static void xen_restart(enum reboot_mode reboot_mode, const char *cmd)
6abb749e SS	198	{
	199	struct sched_shutdown r = { .reason = SHUTDOWN_reboot };
	200	int rc;
	201	rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &r);
a91c7775	202	BUG_ON(rc);
6abb749e SS	203	}
	204
	205	static void xen_power_off(void)
	206	{
	207	struct sched_shutdown r = { .reason = SHUTDOWN_poweroff };
	208	int rc;
	209	rc = HYPERVISOR_sched_op(SCHEDOP_shutdown, &r);
a91c7775	210	BUG_ON(rc);
6abb749e SS	211	}
6abb749e SS	212
8b271d57 JG	213	static int xen_cpu_notification(struct notifier_block *self,
	214	unsigned long action,
	215	void *hcpu)
	216	{
	217	switch (action) {
	218	case CPU_STARTING:
	219	xen_percpu_init();
	220	break;
cb9644bf SS	221	case CPU_DYING:
	222	disable_percpu_irq(xen_events_irq);
	223	break;
8b271d57 JG	224	default:
	225	break;
	226	}
	227
	228	return NOTIFY_OK;
	229	}
	230
	231	static struct notifier_block xen_cpu_notifier = {
	232	.notifier_call = xen_cpu_notification,
	233	};
	234
	235	static irqreturn_t xen_arm_callback(int irq, void *arg)
	236	{
	237	xen_hvm_evtchn_do_upcall();
	238	return IRQ_HANDLED;
	239	}
	240
2e01f166 SS	241	/*
	242	* see Documentation/devicetree/bindings/arm/xen.txt for the
	243	* documentation of the Xen Device Tree format.
	244	*/
b3b52fd8	245	#define GRANT_TABLE_PHYSADDR 0
5882bfef	246	void __init xen_early_init(void)
2e01f166	247	{
2e01f166 SS	248	int len;
	249	const char *s = NULL;
	250	const char *version = NULL;
	251	const char *xen_prefix = "xen,xen-";
	252
5882bfef SS	253	xen_node = of_find_compatible_node(NULL, NULL, "xen,xen");
5882bfef SS	254	if (!xen_node) {
2e01f166	255	pr_debug("No Xen support\n");
5882bfef	256	return;
2e01f166	257	}
5882bfef	258	s = of_get_property(xen_node, "compatible", &len);
2e01f166 SS	259	if (strlen(xen_prefix) + 3 < len &&
	260	!strncmp(xen_prefix, s, strlen(xen_prefix)))
	261	version = s + strlen(xen_prefix);
	262	if (version == NULL) {
	263	pr_debug("Xen version not found\n");
5882bfef	264	return;
81e863c3 JG	265	}
81e863c3 JG	266
5882bfef	267	pr_info("Xen %s support found\n", version);
8b271d57	268
2e01f166 SS	269	xen_domain_type = XEN_HVM_DOMAIN;
2e01f166 SS	270
ef61ee0d	271	xen_setup_features();
5ebc77de	272
ef61ee0d SS	273	if (xen_feature(XENFEAT_dom0))
	274	xen_start_info->flags \|= SIF_INITDOMAIN\|SIF_PRIVILEGED;
	275	else
	276	xen_start_info->flags &= ~(SIF_INITDOMAIN\|SIF_PRIVILEGED);
f1dddd11 AB	277
	278	if (!console_set_on_cmdline && !xen_initial_domain())
	279	add_preferred_console("hvc", 0, NULL);
5882bfef SS	280	}
5882bfef SS	281
d22cbe65 SZ	282	static void __init xen_acpi_guest_init(void)
	283	{
	284	#ifdef CONFIG_ACPI
	285	struct xen_hvm_param a;
	286	int interrupt, trigger, polarity;
	287
	288	a.domid = DOMID_SELF;
	289	a.index = HVM_PARAM_CALLBACK_IRQ;
	290
	291	if (HYPERVISOR_hvm_op(HVMOP_get_param, &a)
	292	\|\| (a.value >> 56) != HVM_PARAM_CALLBACK_TYPE_PPI) {
	293	xen_events_irq = 0;
	294	return;
	295	}
	296
	297	interrupt = a.value & 0xff;
	298	trigger = ((a.value >> 8) & 0x1) ? ACPI_EDGE_SENSITIVE
	299	: ACPI_LEVEL_SENSITIVE;
	300	polarity = ((a.value >> 8) & 0x2) ? ACPI_ACTIVE_LOW
	301	: ACPI_ACTIVE_HIGH;
	302	xen_events_irq = acpi_register_gsi(NULL, interrupt, trigger, polarity);
	303	#endif
	304	}
	305
	306	static void __init xen_dt_guest_init(void)
	307	{
	308	xen_events_irq = irq_of_parse_and_map(xen_node, 0);
	309	}
	310
5882bfef SS	311	static int __init xen_guest_init(void)
	312	{
	313	struct xen_add_to_physmap xatp;
	314	struct shared_info *shared_info_page = NULL;
5882bfef SS	315
	316	if (!xen_domain())
	317	return 0;
	318
d22cbe65 SZ	319	if (!acpi_disabled)
	320	xen_acpi_guest_init();
	321	else
	322	xen_dt_guest_init();
	323
5882bfef SS	324	if (!xen_events_irq) {
	325	pr_err("Xen event channel interrupt not found\n");
	326	return -ENODEV;
	327	}
	328
	329	shared_info_page = (struct shared_info *)get_zeroed_page(GFP_KERNEL);
ef61ee0d	330
2e01f166 SS	331	if (!shared_info_page) {
	332	pr_err("not enough memory\n");
	333	return -ENOMEM;
	334	}
	335	xatp.domid = DOMID_SELF;
	336	xatp.idx = 0;
	337	xatp.space = XENMAPSPACE_shared_info;
250c9af3	338	xatp.gpfn = virt_to_gfn(shared_info_page);
2e01f166 SS	339	if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
	340	BUG();
	341
	342	HYPERVISOR_shared_info = (struct shared_info *)shared_info_page;
	343
	344	/* xen_vcpu is a pointer to the vcpu_info struct in the shared_info
	345	* page, we use it in the event channel upcall and in some pvclock
9a9ab3cc	346	* related functions.
2e01f166 SS	347	* The shared info contains exactly 1 CPU (the boot CPU). The guest
2e01f166 SS	348	* is required to use VCPUOP_register_vcpu_info to place vcpu info
9a9ab3cc SS	349	* for secondary CPUs as they are brought up.
	350	* For uniformity we use VCPUOP_register_vcpu_info even on cpu0.
	351	*/
	352	xen_vcpu_info = __alloc_percpu(sizeof(struct vcpu_info),
	353	sizeof(struct vcpu_info));
	354	if (xen_vcpu_info == NULL)
	355	return -ENOMEM;
b3b52fd8	356
3cf4095d SZ	357	xen_auto_xlat_grant_frames.count = gnttab_max_grant_frames();
	358	if (xen_xlate_map_ballooned_pages(&xen_auto_xlat_grant_frames.pfn,
	359	&xen_auto_xlat_grant_frames.vaddr,
	360	xen_auto_xlat_grant_frames.count)) {
efaf30a3 KRW	361	free_percpu(xen_vcpu_info);
	362	return -ENOMEM;
	363	}
b3b52fd8 SS	364	gnttab_init();
	365	if (!xen_initial_domain())
	366	xenbus_probe(NULL);
	367
e1a9c16b JG	368	/*
	369	* Making sure board specific code will not set up ops for
	370	* cpu idle and cpu freq.
	371	*/
	372	disable_cpuidle();
	373	disable_cpufreq();
	374
8b271d57 JG	375	xen_init_IRQ();
	376
	377	if (request_percpu_irq(xen_events_irq, xen_arm_callback,
	378	"events", &xen_vcpu)) {
	379	pr_err("Error request IRQ %d\n", xen_events_irq);
	380	return -EINVAL;
	381	}
	382
	383	xen_percpu_init();
	384
	385	register_cpu_notifier(&xen_cpu_notifier);
	386
34e38523 SS	387	pv_time_ops.steal_clock = xen_stolen_accounting;
34e38523 SS	388	static_key_slow_inc(&paravirt_steal_enabled);
7d5f6f81 SS	389	if (xen_initial_domain())
7d5f6f81 SS	390	pvclock_gtod_register_notifier(&xen_pvclock_gtod_notifier);
34e38523	391
1aa3d8d9 SS	392	return 0;
1aa3d8d9 SS	393	}
8b271d57	394	early_initcall(xen_guest_init);
1aa3d8d9 SS	395
	396	static int __init xen_pm_init(void)
	397	{
9dd4b294 RH	398	if (!xen_domain())
	399	return -ENODEV;
	400
6abb749e SS	401	pm_power_off = xen_power_off;
6abb749e SS	402	arm_pm_restart = xen_restart;
e709fba1 SS	403	if (!xen_initial_domain()) {
	404	struct timespec64 ts;
	405	xen_read_wallclock(&ts);
	406	do_settimeofday64(&ts);
	407	}
6abb749e	408
2e01f166 SS	409	return 0;
2e01f166 SS	410	}
9dd4b294	411	late_initcall(xen_pm_init);
0ec53ecf	412
79390289 SS	413
	414	/* empty stubs */
	415	void xen_arch_pre_suspend(void) { }
	416	void xen_arch_post_suspend(int suspend_cancelled) { }
	417	void xen_timer_resume(void) { }
	418	void xen_arch_resume(void) { }
ffb7dbed	419	void xen_arch_suspend(void) { }
79390289 SS	420
79390289 SS	421
d5f985c8	422	/* In the hypercall.S file. */
911dec0d KRW	423	EXPORT_SYMBOL_GPL(HYPERVISOR_event_channel_op);
911dec0d KRW	424	EXPORT_SYMBOL_GPL(HYPERVISOR_grant_table_op);
ab277bbf SS	425	EXPORT_SYMBOL_GPL(HYPERVISOR_xen_version);
	426	EXPORT_SYMBOL_GPL(HYPERVISOR_console_io);
	427	EXPORT_SYMBOL_GPL(HYPERVISOR_sched_op);
	428	EXPORT_SYMBOL_GPL(HYPERVISOR_hvm_op);
	429	EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
	430	EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
ea0af613	431	EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
176455e9	432	EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
72d39c69	433	EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
9f1d3414	434	EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
911dec0d	435	EXPORT_SYMBOL_GPL(privcmd_call);