x86/xen/enlighten.c

/*
 * Core of Xen paravirt_ops implementation.
 *
 * This file contains the xen_paravirt_ops structure itself, and the
 * implementations for:
 * - privileged instructions
 * - interrupt flags
 * - segment operations
 * - booting and setup
 *
 * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
 */

#include <linux/cpu.h>
#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/smp.h>
#include <linux/preempt.h>
#include <linux/hardirq.h>
#include <linux/percpu.h>
#include <linux/delay.h>
#include <linux/start_kernel.h>
#include <linux/sched.h>
#include <linux/kprobes.h>
#include <linux/bootmem.h>
#include <linux/export.h>
#include <linux/mm.h>
#include <linux/page-flags.h>
#include <linux/highmem.h>
#include <linux/console.h>
#include <linux/pci.h>
#include <linux/gfp.h>
#include <linux/memblock.h>
#include <linux/edd.h>
#include <linux/frame.h>

#include <linux/kexec.h>

#include <xen/xen.h>
#include <xen/events.h>
#include <xen/interface/xen.h>
#include <xen/interface/version.h>
#include <xen/interface/physdev.h>
#include <xen/interface/vcpu.h>
#include <xen/interface/memory.h>
#include <xen/interface/nmi.h>
#include <xen/interface/xen-mca.h>
#include <xen/interface/hvm/start_info.h>
#include <xen/features.h>
#include <xen/page.h>
#include <xen/hvm.h>
#include <xen/hvc-console.h>
#include <xen/acpi.h>

#include <asm/paravirt.h>
#include <asm/apic.h>
#include <asm/page.h>
#include <asm/xen/pci.h>
#include <asm/xen/hypercall.h>
#include <asm/xen/hypervisor.h>
#include <asm/xen/cpuid.h>
#include <asm/fixmap.h>
#include <asm/processor.h>
#include <asm/proto.h>
#include <asm/msr-index.h>
#include <asm/traps.h>
#include <asm/setup.h>
#include <asm/desc.h>
#include <asm/pgalloc.h>
#include <asm/pgtable.h>
#include <asm/tlbflush.h>
#include <asm/reboot.h>
#include <asm/stackprotector.h>
#include <asm/hypervisor.h>
#include <asm/mach_traps.h>
#include <asm/mwait.h>
#include <asm/pci_x86.h>
#include <asm/cpu.h>
#include <asm/e820/api.h>

#ifdef CONFIG_ACPI
#include <linux/acpi.h>
#include <asm/acpi.h>
#include <acpi/pdc_intel.h>
#include <acpi/processor.h>
#include <xen/interface/platform.h>
#endif

#include "xen-ops.h"
#include "mmu.h"
#include "smp.h"
#include "multicalls.h"
#include "pmu.h"

EXPORT_SYMBOL_GPL(hypercall_page);

/*
 * Pointer to the xen_vcpu_info structure or
 * &HYPERVISOR_shared_info->vcpu_info[cpu]. See xen_hvm_init_shared_info
 * and xen_vcpu_setup for details. By default it points to share_info->vcpu_info
 * but if the hypervisor supports VCPUOP_register_vcpu_info then it can point
 * to xen_vcpu_info. The pointer is used in __xen_evtchn_do_upcall to
 * acknowledge pending events.
 * Also more subtly it is used by the patched version of irq enable/disable
 * e.g. xen_irq_enable_direct and xen_iret in PV mode.
 *
 * The desire to be able to do those mask/unmask operations as a single
 * instruction by using the per-cpu offset held in %gs is the real reason
 * vcpu info is in a per-cpu pointer and the original reason for this
 * hypercall.
 *
 */
DEFINE_PER_CPU(struct vcpu_info *, xen_vcpu);

/*
 * Per CPU pages used if hypervisor supports VCPUOP_register_vcpu_info
 * hypercall. This can be used both in PV and PVHVM mode. The structure
 * overrides the default per_cpu(xen_vcpu, cpu) value.
 */
DEFINE_PER_CPU(struct vcpu_info, xen_vcpu_info);

/* Linux <-> Xen vCPU id mapping */
DEFINE_PER_CPU(uint32_t, xen_vcpu_id);
EXPORT_PER_CPU_SYMBOL(xen_vcpu_id);

enum xen_domain_type xen_domain_type = XEN_NATIVE;
EXPORT_SYMBOL_GPL(xen_domain_type);

unsigned long *machine_to_phys_mapping = (void *)MACH2PHYS_VIRT_START;
EXPORT_SYMBOL(machine_to_phys_mapping);
unsigned long  machine_to_phys_nr;
EXPORT_SYMBOL(machine_to_phys_nr);

struct start_info *xen_start_info;
EXPORT_SYMBOL_GPL(xen_start_info);

struct shared_info xen_dummy_shared_info;

void *xen_initial_gdt;

RESERVE_BRK(shared_info_page_brk, PAGE_SIZE);

static int xen_cpu_up_prepare(unsigned int cpu);
static int xen_cpu_up_online(unsigned int cpu);
static int xen_cpu_dead(unsigned int cpu);

/*
 * Point at some empty memory to start with. We map the real shared_info
 * page as soon as fixmap is up and running.
 */
struct shared_info *HYPERVISOR_shared_info = &xen_dummy_shared_info;

/*
 * Flag to determine whether vcpu info placement is available on all
 * VCPUs.  We assume it is to start with, and then set it to zero on
 * the first failure.  This is because it can succeed on some VCPUs
 * and not others, since it can involve hypervisor memory allocation,
 * or because the guest failed to guarantee all the appropriate
 * constraints on all VCPUs (ie buffer can't cross a page boundary).
 *
 * Note that any particular CPU may be using a placed vcpu structure,
 * but we can only optimise if the all are.
 *
 * 0: not available, 1: available
 */
static int have_vcpu_info_placement = 1;

struct tls_descs {
	struct desc_struct desc[3];
};

/*
 * Updating the 3 TLS descriptors in the GDT on every task switch is
 * surprisingly expensive so we avoid updating them if they haven't
 * changed.  Since Xen writes different descriptors than the one
 * passed in the update_descriptor hypercall we keep shadow copies to
 * compare against.
 */
static DEFINE_PER_CPU(struct tls_descs, shadow_tls_desc);

#ifdef CONFIG_XEN_PVH
/*
 * PVH variables.
 *
 * xen_pvh and pvh_bootparams need to live in data segment since they
 * are used after startup_{32|64}, which clear .bss, are invoked.
 */
bool xen_pvh __attribute__((section(".data"))) = 0;
struct boot_params pvh_bootparams __attribute__((section(".data")));

struct hvm_start_info pvh_start_info;
unsigned int pvh_start_info_sz = sizeof(pvh_start_info);
#endif

static void clamp_max_cpus(void)
{
#ifdef CONFIG_SMP
	if (setup_max_cpus > MAX_VIRT_CPUS)
		setup_max_cpus = MAX_VIRT_CPUS;
#endif
}

void xen_vcpu_setup(int cpu)
{
	struct vcpu_register_vcpu_info info;
	int err;
	struct vcpu_info *vcpup;

	BUG_ON(HYPERVISOR_shared_info == &xen_dummy_shared_info);

	/*
	 * This path is called twice on PVHVM - first during bootup via
	 * smp_init -> xen_hvm_cpu_notify, and then if the VCPU is being
	 * hotplugged: cpu_up -> xen_hvm_cpu_notify.
	 * As we can only do the VCPUOP_register_vcpu_info once lets
	 * not over-write its result.
	 *
	 * For PV it is called during restore (xen_vcpu_restore) and bootup
	 * (xen_setup_vcpu_info_placement). The hotplug mechanism does not
	 * use this function.
	 */
	if (xen_hvm_domain()) {
		if (per_cpu(xen_vcpu, cpu) == &per_cpu(xen_vcpu_info, cpu))
			return;
	}
	if (xen_vcpu_nr(cpu) < MAX_VIRT_CPUS)
		per_cpu(xen_vcpu, cpu) =
			&HYPERVISOR_shared_info->vcpu_info[xen_vcpu_nr(cpu)];

	if (!have_vcpu_info_placement) {
		if (cpu >= MAX_VIRT_CPUS)
			clamp_max_cpus();
		return;
	}

	vcpup = &per_cpu(xen_vcpu_info, cpu);
	info.mfn = arbitrary_virt_to_mfn(vcpup);
	info.offset = offset_in_page(vcpup);

	/* Check to see if the hypervisor will put the vcpu_info
	   structure where we want it, which allows direct access via
	   a percpu-variable.
	   N.B. This hypercall can _only_ be called once per CPU. Subsequent
	   calls will error out with -EINVAL. This is due to the fact that
	   hypervisor has no unregister variant and this hypercall does not
	   allow to over-write info.mfn and info.offset.
	 */
	err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, xen_vcpu_nr(cpu),
				 &info);

	if (err) {
		printk(KERN_DEBUG "register_vcpu_info failed: err=%d\n", err);
		have_vcpu_info_placement = 0;
		clamp_max_cpus();
	} else {
		/* This cpu is using the registered vcpu info, even if
		   later ones fail to. */
		per_cpu(xen_vcpu, cpu) = vcpup;
	}
}

/*
 * On restore, set the vcpu placement up again.
 * If it fails, then we're in a bad state, since
 * we can't back out from using it...
 */
void xen_vcpu_restore(void)
{
	int cpu;

	for_each_possible_cpu(cpu) {
		bool other_cpu = (cpu != smp_processor_id());
		bool is_up = HYPERVISOR_vcpu_op(VCPUOP_is_up, xen_vcpu_nr(cpu),
						NULL);

		if (other_cpu && is_up &&
		    HYPERVISOR_vcpu_op(VCPUOP_down, xen_vcpu_nr(cpu), NULL))
			BUG();

		xen_setup_runstate_info(cpu);

		if (have_vcpu_info_placement)
			xen_vcpu_setup(cpu);

		if (other_cpu && is_up &&
		    HYPERVISOR_vcpu_op(VCPUOP_up, xen_vcpu_nr(cpu), NULL))
			BUG();
	}
}

static void __init xen_banner(void)
{
	unsigned version = HYPERVISOR_xen_version(XENVER_version, NULL);
	struct xen_extraversion extra;
	HYPERVISOR_xen_version(XENVER_extraversion, &extra);

	pr_info("Booting paravirtualized kernel %son %s\n",
		xen_feature(XENFEAT_auto_translated_physmap) ?
			"with PVH extensions " : "", pv_info.name);
	printk(KERN_INFO "Xen version: %d.%d%s%s\n",
	       version >> 16, version & 0xffff, extra.extraversion,
	       xen_feature(XENFEAT_mmu_pt_update_preserve_ad) ? " (preserve-AD)" : "");
}
/* Check if running on Xen version (major, minor) or later */
bool
xen_running_on_version_or_later(unsigned int major, unsigned int minor)
{
	unsigned int version;

	if (!xen_domain())
		return false;

	version = HYPERVISOR_xen_version(XENVER_version, NULL);
	if ((((version >> 16) == major) && ((version & 0xffff) >= minor)) ||
		((version >> 16) > major))
		return true;
	return false;
}

#define CPUID_THERM_POWER_LEAF 6
#define APERFMPERF_PRESENT 0

static __read_mostly unsigned int cpuid_leaf1_edx_mask = ~0;
static __read_mostly unsigned int cpuid_leaf1_ecx_mask = ~0;

static __read_mostly unsigned int cpuid_leaf1_ecx_set_mask;
static __read_mostly unsigned int cpuid_leaf5_ecx_val;
static __read_mostly unsigned int cpuid_leaf5_edx_val;

static void xen_cpuid(unsigned int *ax, unsigned int *bx,
		      unsigned int *cx, unsigned int *dx)
{
	unsigned maskebx = ~0;
	unsigned maskecx = ~0;
	unsigned maskedx = ~0;
	unsigned setecx = 0;
	/*
	 * Mask out inconvenient features, to try and disable as many
	 * unsupported kernel subsystems as possible.
	 */
	switch (*ax) {
	case 1:
		maskecx = cpuid_leaf1_ecx_mask;
		setecx = cpuid_leaf1_ecx_set_mask;
		maskedx = cpuid_leaf1_edx_mask;
		break;

	case CPUID_MWAIT_LEAF:
		/* Synthesize the values.. */
		*ax = 0;
		*bx = 0;
		*cx = cpuid_leaf5_ecx_val;
		*dx = cpuid_leaf5_edx_val;
		return;

	case CPUID_THERM_POWER_LEAF:
		/* Disabling APERFMPERF for kernel usage */
		maskecx = ~(1 << APERFMPERF_PRESENT);
		break;

	case 0xb:
		/* Suppress extended topology stuff */
		maskebx = 0;
		break;
	}

	asm(XEN_EMULATE_PREFIX "cpuid"
		: "=a" (*ax),
		  "=b" (*bx),
		  "=c" (*cx),
		  "=d" (*dx)
		: "0" (*ax), "2" (*cx));

	*bx &= maskebx;
	*cx &= maskecx;
	*cx |= setecx;
	*dx &= maskedx;
}
STACK_FRAME_NON_STANDARD(xen_cpuid); /* XEN_EMULATE_PREFIX */

static bool __init xen_check_mwait(void)
{
#ifdef CONFIG_ACPI
	struct xen_platform_op op = {
		.cmd			= XENPF_set_processor_pminfo,
		.u.set_pminfo.id	= -1,
		.u.set_pminfo.type	= XEN_PM_PDC,
	};
	uint32_t buf[3];
	unsigned int ax, bx, cx, dx;
	unsigned int mwait_mask;

	/* We need to determine whether it is OK to expose the MWAIT
	 * capability to the kernel to harvest deeper than C3 states from ACPI
	 * _CST using the processor_harvest_xen.c module. For this to work, we
	 * need to gather the MWAIT_LEAF values (which the cstate.c code
	 * checks against). The hypervisor won't expose the MWAIT flag because
	 * it would break backwards compatibility; so we will find out directly
	 * from the hardware and hypercall.
	 */
	if (!xen_initial_domain())
		return false;

	/*
	 * When running under platform earlier than Xen4.2, do not expose
	 * mwait, to avoid the risk of loading native acpi pad driver
	 */
	if (!xen_running_on_version_or_later(4, 2))
		return false;

	ax = 1;
	cx = 0;

	native_cpuid(&ax, &bx, &cx, &dx);

	mwait_mask = (1 << (X86_FEATURE_EST % 32)) |
		     (1 << (X86_FEATURE_MWAIT % 32));

	if ((cx & mwait_mask) != mwait_mask)
		return false;

	/* We need to emulate the MWAIT_LEAF and for that we need both
	 * ecx and edx. The hypercall provides only partial information.
	 */

	ax = CPUID_MWAIT_LEAF;
	bx = 0;
	cx = 0;
	dx = 0;

	native_cpuid(&ax, &bx, &cx, &dx);

	/* Ask the Hypervisor whether to clear ACPI_PDC_C_C2C3_FFH. If so,
	 * don't expose MWAIT_LEAF and let ACPI pick the IOPORT version of C3.
	 */
	buf[0] = ACPI_PDC_REVISION_ID;
	buf[1] = 1;
	buf[2] = (ACPI_PDC_C_CAPABILITY_SMP | ACPI_PDC_EST_CAPABILITY_SWSMP);

	set_xen_guest_handle(op.u.set_pminfo.pdc, buf);

	if ((HYPERVISOR_platform_op(&op) == 0) &&
	    (buf[2] & (ACPI_PDC_C_C1_FFH | ACPI_PDC_C_C2C3_FFH))) {
		cpuid_leaf5_ecx_val = cx;
		cpuid_leaf5_edx_val = dx;
	}
	return true;
#else
	return false;
#endif
}
static void __init xen_init_cpuid_mask(void)
{
	unsigned int ax, bx, cx, dx;
	unsigned int xsave_mask;

	cpuid_leaf1_edx_mask =
		~((1 << X86_FEATURE_MTRR) |  /* disable MTRR */
		  (1 << X86_FEATURE_ACC));   /* thermal monitoring */

	if (!xen_initial_domain())
		cpuid_leaf1_edx_mask &=
			~((1 << X86_FEATURE_ACPI));  /* disable ACPI */

	cpuid_leaf1_ecx_mask &= ~(1 << (X86_FEATURE_X2APIC % 32));

	ax = 1;
	cx = 0;
	cpuid(1, &ax, &bx, &cx, &dx);

	xsave_mask =
		(1 << (X86_FEATURE_XSAVE % 32)) |
		(1 << (X86_FEATURE_OSXSAVE % 32));

	/* Xen will set CR4.OSXSAVE if supported and not disabled by force */
	if ((cx & xsave_mask) != xsave_mask)
		cpuid_leaf1_ecx_mask &= ~xsave_mask; /* disable XSAVE & OSXSAVE */
	if (xen_check_mwait())
		cpuid_leaf1_ecx_set_mask = (1 << (X86_FEATURE_MWAIT % 32));
}

static void xen_set_debugreg(int reg, unsigned long val)
{
	HYPERVISOR_set_debugreg(reg, val);
}

static unsigned long xen_get_debugreg(int reg)
{
	return HYPERVISOR_get_debugreg(reg);
}

static void xen_end_context_switch(struct task_struct *next)
{
	xen_mc_flush();
	paravirt_end_context_switch(next);
}

static unsigned long xen_store_tr(void)
{
	return 0;
}

/*
 * Set the page permissions for a particular virtual address.  If the
 * address is a vmalloc mapping (or other non-linear mapping), then
 * find the linear mapping of the page and also set its protections to
 * match.
 */
static void set_aliased_prot(void *v, pgprot_t prot)
{
	int level;
	pte_t *ptep;
	pte_t pte;
	unsigned long pfn;
	struct page *page;
	unsigned char dummy;

	ptep = lookup_address((unsigned long)v, &level);
	BUG_ON(ptep == NULL);

	pfn = pte_pfn(*ptep);
	page = pfn_to_page(pfn);

	pte = pfn_pte(pfn, prot);

	/*
	 * Careful: update_va_mapping() will fail if the virtual address
	 * we're poking isn't populated in the page tables.  We don't
	 * need to worry about the direct map (that's always in the page
	 * tables), but we need to be careful about vmap space.  In
	 * particular, the top level page table can lazily propagate
	 * entries between processes, so if we've switched mms since we
	 * vmapped the target in the first place, we might not have the
	 * top-level page table entry populated.
	 *
	 * We disable preemption because we want the same mm active when
	 * we probe the target and when we issue the hypercall.  We'll
	 * have the same nominal mm, but if we're a kernel thread, lazy
	 * mm dropping could change our pgd.
	 *
	 * Out of an abundance of caution, this uses __get_user() to fault
	 * in the target address just in case there's some obscure case
	 * in which the target address isn't readable.
	 */

	preempt_disable();

	probe_kernel_read(&dummy, v, 1);

	if (HYPERVISOR_update_va_mapping((unsigned long)v, pte, 0))
		BUG();

	if (!PageHighMem(page)) {
		void *av = __va(PFN_PHYS(pfn));

		if (av != v)
			if (HYPERVISOR_update_va_mapping((unsigned long)av, pte, 0))
				BUG();
	} else
		kmap_flush_unused();

	preempt_enable();
}

static void xen_alloc_ldt(struct desc_struct *ldt, unsigned entries)
{
	const unsigned entries_per_page = PAGE_SIZE / LDT_ENTRY_SIZE;
	int i;

	/*
	 * We need to mark the all aliases of the LDT pages RO.  We
	 * don't need to call vm_flush_aliases(), though, since that's
	 * only responsible for flushing aliases out the TLBs, not the
	 * page tables, and Xen will flush the TLB for us if needed.
	 *
	 * To avoid confusing future readers: none of this is necessary
	 * to load the LDT.  The hypervisor only checks this when the
	 * LDT is faulted in due to subsequent descriptor access.
	 */

	for(i = 0; i < entries; i += entries_per_page)
		set_aliased_prot(ldt + i, PAGE_KERNEL_RO);
}

static void xen_free_ldt(struct desc_struct *ldt, unsigned entries)
{
	const unsigned entries_per_page = PAGE_SIZE / LDT_ENTRY_SIZE;
	int i;

	for(i = 0; i < entries; i += entries_per_page)
		set_aliased_prot(ldt + i, PAGE_KERNEL);
}

static void xen_set_ldt(const void *addr, unsigned entries)
{
	struct mmuext_op *op;
	struct multicall_space mcs = xen_mc_entry(sizeof(*op));

	trace_xen_cpu_set_ldt(addr, entries);

	op = mcs.args;
	op->cmd = MMUEXT_SET_LDT;
	op->arg1.linear_addr = (unsigned long)addr;
	op->arg2.nr_ents = entries;

	MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF);

	xen_mc_issue(PARAVIRT_LAZY_CPU);
}

static void xen_load_gdt(const struct desc_ptr *dtr)
{
	unsigned long va = dtr->address;
	unsigned int size = dtr->size + 1;
	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
	unsigned long frames[pages];
	int f;

	/*
	 * A GDT can be up to 64k in size, which corresponds to 8192
	 * 8-byte entries, or 16 4k pages..
	 */

	BUG_ON(size > 65536);
	BUG_ON(va & ~PAGE_MASK);

	for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
		int level;
		pte_t *ptep;
		unsigned long pfn, mfn;
		void *virt;

		/*
		 * The GDT is per-cpu and is in the percpu data area.
		 * That can be virtually mapped, so we need to do a
		 * page-walk to get the underlying MFN for the
		 * hypercall.  The page can also be in the kernel's
		 * linear range, so we need to RO that mapping too.
		 */
		ptep = lookup_address(va, &level);
		BUG_ON(ptep == NULL);

		pfn = pte_pfn(*ptep);
		mfn = pfn_to_mfn(pfn);
		virt = __va(PFN_PHYS(pfn));

		frames[f] = mfn;

		make_lowmem_page_readonly((void *)va);
		make_lowmem_page_readonly(virt);
	}

	if (HYPERVISOR_set_gdt(frames, size / sizeof(struct desc_struct)))
		BUG();
}

/*
 * load_gdt for early boot, when the gdt is only mapped once
 */
static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
{
	unsigned long va = dtr->address;
	unsigned int size = dtr->size + 1;
	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
	unsigned long frames[pages];
	int f;

	/*
	 * A GDT can be up to 64k in size, which corresponds to 8192
	 * 8-byte entries, or 16 4k pages..
	 */

	BUG_ON(size > 65536);
	BUG_ON(va & ~PAGE_MASK);

	for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
		pte_t pte;
		unsigned long pfn, mfn;

		pfn = virt_to_pfn(va);
		mfn = pfn_to_mfn(pfn);

		pte = pfn_pte(pfn, PAGE_KERNEL_RO);

		if (HYPERVISOR_update_va_mapping((unsigned long)va, pte, 0))
			BUG();

		frames[f] = mfn;
	}

	if (HYPERVISOR_set_gdt(frames, size / sizeof(struct desc_struct)))
		BUG();
}

static inline bool desc_equal(const struct desc_struct *d1,
			      const struct desc_struct *d2)
{
	return d1->a == d2->a && d1->b == d2->b;
}

static void load_TLS_descriptor(struct thread_struct *t,
				unsigned int cpu, unsigned int i)
{
	struct desc_struct *shadow = &per_cpu(shadow_tls_desc, cpu).desc[i];
	struct desc_struct *gdt;
	xmaddr_t maddr;
	struct multicall_space mc;

	if (desc_equal(shadow, &t->tls_array[i]))
		return;

	*shadow = t->tls_array[i];

	gdt = get_cpu_gdt_rw(cpu);
	maddr = arbitrary_virt_to_machine(&gdt[GDT_ENTRY_TLS_MIN+i]);
	mc = __xen_mc_entry(0);

	MULTI_update_descriptor(mc.mc, maddr.maddr, t->tls_array[i]);
}

static void xen_load_tls(struct thread_struct *t, unsigned int cpu)
{
	/*
	 * XXX sleazy hack: If we're being called in a lazy-cpu zone
	 * and lazy gs handling is enabled, it means we're in a
	 * context switch, and %gs has just been saved.  This means we
	 * can zero it out to prevent faults on exit from the
	 * hypervisor if the next process has no %gs.  Either way, it
	 * has been saved, and the new value will get loaded properly.
	 * This will go away as soon as Xen has been modified to not
	 * save/restore %gs for normal hypercalls.
	 *
	 * On x86_64, this hack is not used for %gs, because gs points
	 * to KERNEL_GS_BASE (and uses it for PDA references), so we
	 * must not zero %gs on x86_64
	 *
	 * For x86_64, we need to zero %fs, otherwise we may get an
	 * exception between the new %fs descriptor being loaded and
	 * %fs being effectively cleared at __switch_to().
	 */
	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_CPU) {
#ifdef CONFIG_X86_32
		lazy_load_gs(0);
#else
		loadsegment(fs, 0);
#endif
	}

	xen_mc_batch();

	load_TLS_descriptor(t, cpu, 0);
	load_TLS_descriptor(t, cpu, 1);
	load_TLS_descriptor(t, cpu, 2);

	xen_mc_issue(PARAVIRT_LAZY_CPU);
}

#ifdef CONFIG_X86_64
static void xen_load_gs_index(unsigned int idx)
{
	if (HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, idx))
		BUG();
}
#endif

static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum,
				const void *ptr)
{
	xmaddr_t mach_lp = arbitrary_virt_to_machine(&dt[entrynum]);
	u64 entry = *(u64 *)ptr;

	trace_xen_cpu_write_ldt_entry(dt, entrynum, entry);

	preempt_disable();

	xen_mc_flush();
	if (HYPERVISOR_update_descriptor(mach_lp.maddr, entry))
		BUG();

	preempt_enable();
}

static int cvt_gate_to_trap(int vector, const gate_desc *val,
			    struct trap_info *info)
{
	unsigned long addr;

	if (val->type != GATE_TRAP && val->type != GATE_INTERRUPT)
		return 0;

	info->vector = vector;

	addr = gate_offset(*val);
#ifdef CONFIG_X86_64
	/*
	 * Look for known traps using IST, and substitute them
	 * appropriately.  The debugger ones are the only ones we care
	 * about.  Xen will handle faults like double_fault,
	 * so we should never see them.  Warn if
	 * there's an unexpected IST-using fault handler.
	 */
	if (addr == (unsigned long)debug)
		addr = (unsigned long)xen_debug;
	else if (addr == (unsigned long)int3)
		addr = (unsigned long)xen_int3;
	else if (addr == (unsigned long)stack_segment)
		addr = (unsigned long)xen_stack_segment;
	else if (addr == (unsigned long)double_fault) {
		/* Don't need to handle these */
		return 0;
#ifdef CONFIG_X86_MCE
	} else if (addr == (unsigned long)machine_check) {
		/*
		 * when xen hypervisor inject vMCE to guest,
		 * use native mce handler to handle it
		 */
		;
#endif
	} else if (addr == (unsigned long)nmi)
		/*
		 * Use the native version as well.
		 */
		;
	else {
		/* Some other trap using IST? */
		if (WARN_ON(val->ist != 0))
			return 0;
	}
#endif	/* CONFIG_X86_64 */
	info->address = addr;

	info->cs = gate_segment(*val);
	info->flags = val->dpl;
	/* interrupt gates clear IF */
	if (val->type == GATE_INTERRUPT)
		info->flags |= 1 << 2;

	return 1;
}

/* Locations of each CPU's IDT */
static DEFINE_PER_CPU(struct desc_ptr, idt_desc);

/* Set an IDT entry.  If the entry is part of the current IDT, then
   also update Xen. */
static void xen_write_idt_entry(gate_desc *dt, int entrynum, const gate_desc *g)
{
	unsigned long p = (unsigned long)&dt[entrynum];
	unsigned long start, end;

	trace_xen_cpu_write_idt_entry(dt, entrynum, g);

	preempt_disable();

	start = __this_cpu_read(idt_desc.address);
	end = start + __this_cpu_read(idt_desc.size) + 1;

	xen_mc_flush();

	native_write_idt_entry(dt, entrynum, g);

	if (p >= start && (p + 8) <= end) {
		struct trap_info info[2];

		info[1].address = 0;

		if (cvt_gate_to_trap(entrynum, g, &info[0]))
			if (HYPERVISOR_set_trap_table(info))
				BUG();
	}

	preempt_enable();
}

static void xen_convert_trap_info(const struct desc_ptr *desc,
				  struct trap_info *traps)
{
	unsigned in, out, count;

	count = (desc->size+1) / sizeof(gate_desc);
	BUG_ON(count > 256);

	for (in = out = 0; in < count; in++) {
		gate_desc *entry = (gate_desc*)(desc->address) + in;

		if (cvt_gate_to_trap(in, entry, &traps[out]))
			out++;
	}
	traps[out].address = 0;
}

void xen_copy_trap_info(struct trap_info *traps)
{
	const struct desc_ptr *desc = this_cpu_ptr(&idt_desc);

	xen_convert_trap_info(desc, traps);
}

/* Load a new IDT into Xen.  In principle this can be per-CPU, so we
   hold a spinlock to protect the static traps[] array (static because
   it avoids allocation, and saves stack space). */
static void xen_load_idt(const struct desc_ptr *desc)
{
	static DEFINE_SPINLOCK(lock);
	static struct trap_info traps[257];

	trace_xen_cpu_load_idt(desc);

	spin_lock(&lock);

	memcpy(this_cpu_ptr(&idt_desc), desc, sizeof(idt_desc));

	xen_convert_trap_info(desc, traps);

	xen_mc_flush();
	if (HYPERVISOR_set_trap_table(traps))
		BUG();

	spin_unlock(&lock);
}

/* Write a GDT descriptor entry.  Ignore LDT descriptors, since
   they're handled differently. */
static void xen_write_gdt_entry(struct desc_struct *dt, int entry,
				const void *desc, int type)
{
	trace_xen_cpu_write_gdt_entry(dt, entry, desc, type);

	preempt_disable();

	switch (type) {
	case DESC_LDT:
	case DESC_TSS:
		/* ignore */
		break;

	default: {
		xmaddr_t maddr = arbitrary_virt_to_machine(&dt[entry]);

		xen_mc_flush();
		if (HYPERVISOR_update_descriptor(maddr.maddr, *(u64 *)desc))
			BUG();
	}

	}

	preempt_enable();
}

/*
 * Version of write_gdt_entry for use at early boot-time needed to
 * update an entry as simply as possible.
 */
static void __init xen_write_gdt_entry_boot(struct desc_struct *dt, int entry,
					    const void *desc, int type)
{
	trace_xen_cpu_write_gdt_entry(dt, entry, desc, type);

	switch (type) {
	case DESC_LDT:
	case DESC_TSS:
		/* ignore */
		break;

	default: {
		xmaddr_t maddr = virt_to_machine(&dt[entry]);

		if (HYPERVISOR_update_descriptor(maddr.maddr, *(u64 *)desc))
			dt[entry] = *(struct desc_struct *)desc;
	}

	}
}

static void xen_load_sp0(struct tss_struct *tss,
			 struct thread_struct *thread)
{
	struct multicall_space mcs;

	mcs = xen_mc_entry(0);
	MULTI_stack_switch(mcs.mc, __KERNEL_DS, thread->sp0);
	xen_mc_issue(PARAVIRT_LAZY_CPU);
	tss->x86_tss.sp0 = thread->sp0;
}

void xen_set_iopl_mask(unsigned mask)
{
	struct physdev_set_iopl set_iopl;

	/* Force the change at ring 0. */
	set_iopl.iopl = (mask == 0) ? 1 : (mask >> 12) & 3;
	HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
}

static void xen_io_delay(void)
{
}

static DEFINE_PER_CPU(unsigned long, xen_cr0_value);

static unsigned long xen_read_cr0(void)
{
	unsigned long cr0 = this_cpu_read(xen_cr0_value);

	if (unlikely(cr0 == 0)) {
		cr0 = native_read_cr0();
		this_cpu_write(xen_cr0_value, cr0);
	}

	return cr0;
}

static void xen_write_cr0(unsigned long cr0)
{
	struct multicall_space mcs;

	this_cpu_write(xen_cr0_value, cr0);

	/* Only pay attention to cr0.TS; everything else is
	   ignored. */
	mcs = xen_mc_entry(0);

	MULTI_fpu_taskswitch(mcs.mc, (cr0 & X86_CR0_TS) != 0);

	xen_mc_issue(PARAVIRT_LAZY_CPU);
}

static void xen_write_cr4(unsigned long cr4)
{
	cr4 &= ~(X86_CR4_PGE | X86_CR4_PSE | X86_CR4_PCE);

	native_write_cr4(cr4);
}
#ifdef CONFIG_X86_64
static inline unsigned long xen_read_cr8(void)
{
	return 0;
}
static inline void xen_write_cr8(unsigned long val)
{
	BUG_ON(val);
}
#endif

static u64 xen_read_msr_safe(unsigned int msr, int *err)
{
	u64 val;

	if (pmu_msr_read(msr, &val, err))
		return val;

	val = native_read_msr_safe(msr, err);
	switch (msr) {
	case MSR_IA32_APICBASE:
#ifdef CONFIG_X86_X2APIC
		if (!(cpuid_ecx(1) & (1 << (X86_FEATURE_X2APIC & 31))))
#endif
			val &= ~X2APIC_ENABLE;
		break;
	}
	return val;
}

static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high)
{
	int ret;

	ret = 0;

	switch (msr) {
#ifdef CONFIG_X86_64
		unsigned which;
		u64 base;

	case MSR_FS_BASE:		which = SEGBASE_FS; goto set;
	case MSR_KERNEL_GS_BASE:	which = SEGBASE_GS_USER; goto set;
	case MSR_GS_BASE:		which = SEGBASE_GS_KERNEL; goto set;

	set:
		base = ((u64)high << 32) | low;
		if (HYPERVISOR_set_segment_base(which, base) != 0)
			ret = -EIO;
		break;
#endif

	case MSR_STAR:
	case MSR_CSTAR:
	case MSR_LSTAR:
	case MSR_SYSCALL_MASK:
	case MSR_IA32_SYSENTER_CS:
	case MSR_IA32_SYSENTER_ESP:
	case MSR_IA32_SYSENTER_EIP:
		/* Fast syscall setup is all done in hypercalls, so
		   these are all ignored.  Stub them out here to stop
		   Xen console noise. */
		break;

	default:
		if (!pmu_msr_write(msr, low, high, &ret))
			ret = native_write_msr_safe(msr, low, high);
	}

	return ret;
}

static u64 xen_read_msr(unsigned int msr)
{
	/*
	 * This will silently swallow a #GP from RDMSR.  It may be worth
	 * changing that.
	 */
	int err;

	return xen_read_msr_safe(msr, &err);
}

static void xen_write_msr(unsigned int msr, unsigned low, unsigned high)
{
	/*
	 * This will silently swallow a #GP from WRMSR.  It may be worth
	 * changing that.
	 */
	xen_write_msr_safe(msr, low, high);
}

void xen_setup_shared_info(void)
{
	if (!xen_feature(XENFEAT_auto_translated_physmap)) {
		set_fixmap(FIX_PARAVIRT_BOOTMAP,
			   xen_start_info->shared_info);

		HYPERVISOR_shared_info =
			(struct shared_info *)fix_to_virt(FIX_PARAVIRT_BOOTMAP);
	} else
		HYPERVISOR_shared_info =
			(struct shared_info *)__va(xen_start_info->shared_info);

#ifndef CONFIG_SMP
	/* In UP this is as good a place as any to set up shared info */
	xen_setup_vcpu_info_placement();
#endif

	xen_setup_mfn_list_list();
}

/* This is called once we have the cpu_possible_mask */
void xen_setup_vcpu_info_placement(void)
{
	int cpu;

	for_each_possible_cpu(cpu) {
		/* Set up direct vCPU id mapping for PV guests. */
		per_cpu(xen_vcpu_id, cpu) = cpu;
		xen_vcpu_setup(cpu);
	}

	/*
	 * xen_vcpu_setup managed to place the vcpu_info within the
	 * percpu area for all cpus, so make use of it.
	 */
	if (have_vcpu_info_placement) {
		pv_irq_ops.save_fl = __PV_IS_CALLEE_SAVE(xen_save_fl_direct);
		pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(xen_restore_fl_direct);
		pv_irq_ops.irq_disable = __PV_IS_CALLEE_SAVE(xen_irq_disable_direct);
		pv_irq_ops.irq_enable = __PV_IS_CALLEE_SAVE(xen_irq_enable_direct);
		pv_mmu_ops.read_cr2 = xen_read_cr2_direct;
	}
}

static unsigned xen_patch(u8 type, u16 clobbers, void *insnbuf,
			  unsigned long addr, unsigned len)
{
	char *start, *end, *reloc;
	unsigned ret;

	start = end = reloc = NULL;

#define SITE(op, x)							\
	case PARAVIRT_PATCH(op.x):					\
	if (have_vcpu_info_placement) {					\
		start = (char *)xen_##x##_direct;			\
		end = xen_##x##_direct_end;				\
		reloc = xen_##x##_direct_reloc;				\
	}								\
	goto patch_site

	switch (type) {
		SITE(pv_irq_ops, irq_enable);
		SITE(pv_irq_ops, irq_disable);
		SITE(pv_irq_ops, save_fl);
		SITE(pv_irq_ops, restore_fl);
#undef SITE

	patch_site:
		if (start == NULL || (end-start) > len)
			goto default_patch;

		ret = paravirt_patch_insns(insnbuf, len, start, end);

		/* Note: because reloc is assigned from something that
		   appears to be an array, gcc assumes it's non-null,
		   but doesn't know its relationship with start and
		   end. */
		if (reloc > start && reloc < end) {
			int reloc_off = reloc - start;
			long *relocp = (long *)(insnbuf + reloc_off);
			long delta = start - (char *)addr;

			*relocp += delta;
		}
		break;

	default_patch:
	default:
		ret = paravirt_patch_default(type, clobbers, insnbuf,
					     addr, len);
		break;
	}

	return ret;
}

static const struct pv_info xen_info __initconst = {
	.shared_kernel_pmd = 0,

#ifdef CONFIG_X86_64
	.extra_user_64bit_cs = FLAT_USER_CS64,
#endif
	.name = "Xen",
};

static const struct pv_init_ops xen_init_ops __initconst = {
	.patch = xen_patch,
};

static const struct pv_cpu_ops xen_cpu_ops __initconst = {
	.cpuid = xen_cpuid,

	.set_debugreg = xen_set_debugreg,
	.get_debugreg = xen_get_debugreg,

	.read_cr0 = xen_read_cr0,
	.write_cr0 = xen_write_cr0,

	.read_cr4 = native_read_cr4,
	.write_cr4 = xen_write_cr4,

#ifdef CONFIG_X86_64
	.read_cr8 = xen_read_cr8,
	.write_cr8 = xen_write_cr8,
#endif

	.wbinvd = native_wbinvd,

	.read_msr = xen_read_msr,
	.write_msr = xen_write_msr,

	.read_msr_safe = xen_read_msr_safe,
	.write_msr_safe = xen_write_msr_safe,

	.read_pmc = xen_read_pmc,

	.iret = xen_iret,
#ifdef CONFIG_X86_64
	.usergs_sysret64 = xen_sysret64,
#endif

	.load_tr_desc = paravirt_nop,
	.set_ldt = xen_set_ldt,
	.load_gdt = xen_load_gdt,
	.load_idt = xen_load_idt,
	.load_tls = xen_load_tls,
#ifdef CONFIG_X86_64
	.load_gs_index = xen_load_gs_index,
#endif

	.alloc_ldt = xen_alloc_ldt,
	.free_ldt = xen_free_ldt,

	.store_idt = native_store_idt,
	.store_tr = xen_store_tr,

	.write_ldt_entry = xen_write_ldt_entry,
	.write_gdt_entry = xen_write_gdt_entry,
	.write_idt_entry = xen_write_idt_entry,
	.load_sp0 = xen_load_sp0,

	.set_iopl_mask = xen_set_iopl_mask,
	.io_delay = xen_io_delay,

	/* Xen takes care of %gs when switching to usermode for us */
	.swapgs = paravirt_nop,

	.start_context_switch = paravirt_start_context_switch,
	.end_context_switch = xen_end_context_switch,
};

static void xen_reboot(int reason)
{
	struct sched_shutdown r = { .reason = reason };
	int cpu;

	for_each_online_cpu(cpu)
		xen_pmu_finish(cpu);

	if (HYPERVISOR_sched_op(SCHEDOP_shutdown, &r))
		BUG();
}

static void xen_restart(char *msg)
{
	xen_reboot(SHUTDOWN_reboot);
}

static void xen_emergency_restart(void)
{
	xen_reboot(SHUTDOWN_reboot);
}

static void xen_machine_halt(void)
{
	xen_reboot(SHUTDOWN_poweroff);
}

static void xen_machine_power_off(void)
{
	if (pm_power_off)
		pm_power_off();
	xen_reboot(SHUTDOWN_poweroff);
}

static void xen_crash_shutdown(struct pt_regs *regs)
{
	xen_reboot(SHUTDOWN_crash);
}

static int
xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
{
	if (!kexec_crash_loaded())
		xen_reboot(SHUTDOWN_crash);
	return NOTIFY_DONE;
}

static struct notifier_block xen_panic_block = {
	.notifier_call= xen_panic_event,
	.priority = INT_MIN
};

int xen_panic_handler_init(void)
{
	atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
	return 0;
}

static const struct machine_ops xen_machine_ops __initconst = {
	.restart = xen_restart,
	.halt = xen_machine_halt,
	.power_off = xen_machine_power_off,
	.shutdown = xen_machine_halt,
	.crash_shutdown = xen_crash_shutdown,
	.emergency_restart = xen_emergency_restart,
};

static unsigned char xen_get_nmi_reason(void)
{
	unsigned char reason = 0;

	/* Construct a value which looks like it came from port 0x61. */
	if (test_bit(_XEN_NMIREASON_io_error,
		     &HYPERVISOR_shared_info->arch.nmi_reason))
		reason |= NMI_REASON_IOCHK;
	if (test_bit(_XEN_NMIREASON_pci_serr,
		     &HYPERVISOR_shared_info->arch.nmi_reason))
		reason |= NMI_REASON_SERR;

	return reason;
}

static void __init xen_boot_params_init_edd(void)
{
#if IS_ENABLED(CONFIG_EDD)
	struct xen_platform_op op;
	struct edd_info *edd_info;
	u32 *mbr_signature;
	unsigned nr;
	int ret;

	edd_info = boot_params.eddbuf;
	mbr_signature = boot_params.edd_mbr_sig_buffer;

	op.cmd = XENPF_firmware_info;

	op.u.firmware_info.type = XEN_FW_DISK_INFO;
	for (nr = 0; nr < EDDMAXNR; nr++) {
		struct edd_info *info = edd_info + nr;

		op.u.firmware_info.index = nr;
		info->params.length = sizeof(info->params);
		set_xen_guest_handle(op.u.firmware_info.u.disk_info.edd_params,
				     &info->params);
		ret = HYPERVISOR_platform_op(&op);
		if (ret)
			break;

#define C(x) info->x = op.u.firmware_info.u.disk_info.x
		C(device);
		C(version);
		C(interface_support);
		C(legacy_max_cylinder);
		C(legacy_max_head);
		C(legacy_sectors_per_track);
#undef C
	}
	boot_params.eddbuf_entries = nr;

	op.u.firmware_info.type = XEN_FW_DISK_MBR_SIGNATURE;
	for (nr = 0; nr < EDD_MBR_SIG_MAX; nr++) {
		op.u.firmware_info.index = nr;
		ret = HYPERVISOR_platform_op(&op);
		if (ret)
			break;
		mbr_signature[nr] = op.u.firmware_info.u.disk_mbr_signature.mbr_signature;
	}
	boot_params.edd_mbr_sig_buf_entries = nr;
#endif
}

/*
 * Set up the GDT and segment registers for -fstack-protector.  Until
 * we do this, we have to be careful not to call any stack-protected
 * function, which is most of the kernel.
 */
static void xen_setup_gdt(int cpu)
{
	pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry_boot;
	pv_cpu_ops.load_gdt = xen_load_gdt_boot;

	setup_stack_canary_segment(0);
	switch_to_new_gdt(0);

	pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry;
	pv_cpu_ops.load_gdt = xen_load_gdt;
}

static void __init xen_dom0_set_legacy_features(void)
{
	x86_platform.legacy.rtc = 1;
}

static int xen_cpuhp_setup(void)
{
	int rc;

	rc = cpuhp_setup_state_nocalls(CPUHP_XEN_PREPARE,
				       "x86/xen/hvm_guest:prepare",
				       xen_cpu_up_prepare, xen_cpu_dead);
	if (rc >= 0) {
		rc = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
					       "x86/xen/hvm_guest:online",
					       xen_cpu_up_online, NULL);
		if (rc < 0)
			cpuhp_remove_state_nocalls(CPUHP_XEN_PREPARE);
	}

	return rc >= 0 ? 0 : rc;
}

/* First C function to be called on Xen boot */
asmlinkage __visible void __init xen_start_kernel(void)
{
	struct physdev_set_iopl set_iopl;
	unsigned long initrd_start = 0;
	int rc;

	if (!xen_start_info)
		return;

	xen_domain_type = XEN_PV_DOMAIN;

	xen_setup_features();

	xen_setup_machphys_mapping();

	/* Install Xen paravirt ops */
	pv_info = xen_info;
	pv_init_ops = xen_init_ops;
	pv_cpu_ops = xen_cpu_ops;

	x86_platform.get_nmi_reason = xen_get_nmi_reason;

	x86_init.resources.memory_setup = xen_memory_setup;
	x86_init.oem.arch_setup = xen_arch_setup;
	x86_init.oem.banner = xen_banner;

	xen_init_time_ops();

	/*
	 * Set up some pagetable state before starting to set any ptes.
	 */

	xen_init_mmu_ops();

	/* Prevent unwanted bits from being set in PTEs. */
	__supported_pte_mask &= ~_PAGE_GLOBAL;

	/*
	 * Prevent page tables from being allocated in highmem, even
	 * if CONFIG_HIGHPTE is enabled.
	 */
	__userpte_alloc_gfp &= ~__GFP_HIGHMEM;

	/* Work out if we support NX */
	x86_configure_nx();

	/* Get mfn list */
	xen_build_dynamic_phys_to_machine();

	/*
	 * Set up kernel GDT and segment registers, mainly so that
	 * -fstack-protector code can be executed.
	 */
	xen_setup_gdt(0);

	xen_init_irq_ops();
	xen_init_cpuid_mask();

#ifdef CONFIG_X86_LOCAL_APIC
	/*
	 * set up the basic apic ops.
	 */
	xen_init_apic();
#endif

	if (xen_feature(XENFEAT_mmu_pt_update_preserve_ad)) {
		pv_mmu_ops.ptep_modify_prot_start = xen_ptep_modify_prot_start;
		pv_mmu_ops.ptep_modify_prot_commit = xen_ptep_modify_prot_commit;
	}

	machine_ops = xen_machine_ops;

	/*
	 * The only reliable way to retain the initial address of the
	 * percpu gdt_page is to remember it here, so we can go and
	 * mark it RW later, when the initial percpu area is freed.
	 */
	xen_initial_gdt = &per_cpu(gdt_page, 0);

	xen_smp_init();

#ifdef CONFIG_ACPI_NUMA
	/*
	 * The pages we from Xen are not related to machine pages, so
	 * any NUMA information the kernel tries to get from ACPI will
	 * be meaningless.  Prevent it from trying.
	 */
	acpi_numa = -1;
#endif
	/* Don't do the full vcpu_info placement stuff until we have a
	   possible map and a non-dummy shared_info. */
	per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];

	WARN_ON(xen_cpuhp_setup());

	local_irq_disable();
	early_boot_irqs_disabled = true;

	xen_raw_console_write("mapping kernel into physical memory\n");
	xen_setup_kernel_pagetable((pgd_t *)xen_start_info->pt_base,
				   xen_start_info->nr_pages);
	xen_reserve_special_pages();

	/* keep using Xen gdt for now; no urgent need to change it */

#ifdef CONFIG_X86_32
	pv_info.kernel_rpl = 1;
	if (xen_feature(XENFEAT_supervisor_mode_kernel))
		pv_info.kernel_rpl = 0;
#else
	pv_info.kernel_rpl = 0;
#endif
	/* set the limit of our address space */
	xen_reserve_top();

	/*
	 * We used to do this in xen_arch_setup, but that is too late
	 * on AMD were early_cpu_init (run before ->arch_setup()) calls
	 * early_amd_init which pokes 0xcf8 port.
	 */
	set_iopl.iopl = 1;
	rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
	if (rc != 0)
		xen_raw_printk("physdev_op failed %d\n", rc);

#ifdef CONFIG_X86_32
	/* set up basic CPUID stuff */
	cpu_detect(&new_cpu_data);
	set_cpu_cap(&new_cpu_data, X86_FEATURE_FPU);
	new_cpu_data.x86_capability[CPUID_1_EDX] = cpuid_edx(1);
#endif

	if (xen_start_info->mod_start) {
	    if (xen_start_info->flags & SIF_MOD_START_PFN)
		initrd_start = PFN_PHYS(xen_start_info->mod_start);
	    else
		initrd_start = __pa(xen_start_info->mod_start);
	}

	/* Poke various useful things into boot_params */
	boot_params.hdr.type_of_loader = (9 << 4) | 0;
	boot_params.hdr.ramdisk_image = initrd_start;
	boot_params.hdr.ramdisk_size = xen_start_info->mod_len;
	boot_params.hdr.cmd_line_ptr = __pa(xen_start_info->cmd_line);
	boot_params.hdr.hardware_subarch = X86_SUBARCH_XEN;

	if (!xen_initial_domain()) {
		add_preferred_console("xenboot", 0, NULL);
		add_preferred_console("tty", 0, NULL);
		add_preferred_console("hvc", 0, NULL);
		if (pci_xen)
			x86_init.pci.arch_init = pci_xen_init;
	} else {
		const struct dom0_vga_console_info *info =
			(void *)((char *)xen_start_info +
				 xen_start_info->console.dom0.info_off);
		struct xen_platform_op op = {
			.cmd = XENPF_firmware_info,
			.interface_version = XENPF_INTERFACE_VERSION,
			.u.firmware_info.type = XEN_FW_KBD_SHIFT_FLAGS,
		};

		x86_platform.set_legacy_features =
				xen_dom0_set_legacy_features;
		xen_init_vga(info, xen_start_info->console.dom0.info_size);
		xen_start_info->console.domU.mfn = 0;
		xen_start_info->console.domU.evtchn = 0;

		if (HYPERVISOR_platform_op(&op) == 0)
			boot_params.kbd_status = op.u.firmware_info.u.kbd_shift_flags;

		/* Make sure ACS will be enabled */
		pci_request_acs();

		xen_acpi_sleep_register();

		/* Avoid searching for BIOS MP tables */
		x86_init.mpparse.find_smp_config = x86_init_noop;
		x86_init.mpparse.get_smp_config = x86_init_uint_noop;

		xen_boot_params_init_edd();
	}
#ifdef CONFIG_PCI
	/* PCI BIOS service won't work from a PV guest. */
	pci_probe &= ~PCI_PROBE_BIOS;
#endif
	xen_raw_console_write("about to get started...\n");

	/* Let's presume PV guests always boot on vCPU with id 0. */
	per_cpu(xen_vcpu_id, 0) = 0;

	xen_setup_runstate_info(0);

	xen_efi_init();

	/* Start the world */
#ifdef CONFIG_X86_32
	i386_start_kernel();
#else
	cr4_init_shadow(); /* 32b kernel does this in i386_start_kernel() */
	x86_64_start_reservations((char *)__pa_symbol(&boot_params));
#endif
}

#ifdef CONFIG_XEN_PVH

static void xen_pvh_arch_setup(void)
{
#ifdef CONFIG_ACPI
	/* Make sure we don't fall back to (default) ACPI_IRQ_MODEL_PIC. */
	if (nr_ioapics == 0)
		acpi_irq_model = ACPI_IRQ_MODEL_PLATFORM;
#endif
}

static void __init init_pvh_bootparams(void)
{
	struct xen_memory_map memmap;
	unsigned int i;
	int rc;

	memset(&pvh_bootparams, 0, sizeof(pvh_bootparams));

	memmap.nr_entries = ARRAY_SIZE(pvh_bootparams.e820_table);
	set_xen_guest_handle(memmap.buffer, pvh_bootparams.e820_table);
	rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
	if (rc) {
		xen_raw_printk("XENMEM_memory_map failed (%d)\n", rc);
		BUG();
	}

	if (memmap.nr_entries < E820_MAX_ENTRIES_ZEROPAGE - 1) {
		pvh_bootparams.e820_table[memmap.nr_entries].addr =
			ISA_START_ADDRESS;
		pvh_bootparams.e820_table[memmap.nr_entries].size =
			ISA_END_ADDRESS - ISA_START_ADDRESS;
		pvh_bootparams.e820_table[memmap.nr_entries].type =
			E820_TYPE_RESERVED;
		memmap.nr_entries++;
	} else
		xen_raw_printk("Warning: Can fit ISA range into e820\n");

	pvh_bootparams.e820_entries = memmap.nr_entries;
	for (i = 0; i < pvh_bootparams.e820_entries; i++)
		e820__range_add(pvh_bootparams.e820_table[i].addr,
				pvh_bootparams.e820_table[i].size,
				pvh_bootparams.e820_table[i].type);

	e820__update_table(e820_table);

	pvh_bootparams.hdr.cmd_line_ptr =
		pvh_start_info.cmdline_paddr;

	/* The first module is always ramdisk. */
	if (pvh_start_info.nr_modules) {
		struct hvm_modlist_entry *modaddr =
			__va(pvh_start_info.modlist_paddr);
		pvh_bootparams.hdr.ramdisk_image = modaddr->paddr;
		pvh_bootparams.hdr.ramdisk_size = modaddr->size;
	}

	/*
	 * See Documentation/x86/boot.txt.
	 *
	 * Version 2.12 supports Xen entry point but we will use default x86/PC
	 * environment (i.e. hardware_subarch 0).
	 */
	pvh_bootparams.hdr.version = 0x212;
	pvh_bootparams.hdr.type_of_loader = (9 << 4) | 0; /* Xen loader */
}

/*
 * This routine (and those that it might call) should not use
 * anything that lives in .bss since that segment will be cleared later.
 */
void __init xen_prepare_pvh(void)
{
	u32 msr;
	u64 pfn;

	if (pvh_start_info.magic != XEN_HVM_START_MAGIC_VALUE) {
		xen_raw_printk("Error: Unexpected magic value (0x%08x)\n",
				pvh_start_info.magic);
		BUG();
	}

	xen_pvh = 1;

	msr = cpuid_ebx(xen_cpuid_base() + 2);
	pfn = __pa(hypercall_page);
	wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));

	init_pvh_bootparams();

	x86_init.oem.arch_setup = xen_pvh_arch_setup;
}
#endif

void __ref xen_hvm_init_shared_info(void)
{
	int cpu;
	struct xen_add_to_physmap xatp;
	static struct shared_info *shared_info_page = 0;

	if (!shared_info_page)
		shared_info_page = (struct shared_info *)
			extend_brk(PAGE_SIZE, PAGE_SIZE);
	xatp.domid = DOMID_SELF;
	xatp.idx = 0;
	xatp.space = XENMAPSPACE_shared_info;
	xatp.gpfn = __pa(shared_info_page) >> PAGE_SHIFT;
	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. We don't need the vcpu_info placement
	 * optimizations because we don't use any pv_mmu or pv_irq op on
	 * HVM.
	 * When xen_hvm_init_shared_info is run at boot time only vcpu 0 is
	 * online but xen_hvm_init_shared_info is run at resume time too and
	 * in that case multiple vcpus might be online. */
	for_each_online_cpu(cpu) {
		/* Leave it to be NULL. */
		if (xen_vcpu_nr(cpu) >= MAX_VIRT_CPUS)
			continue;
		per_cpu(xen_vcpu, cpu) =
			&HYPERVISOR_shared_info->vcpu_info[xen_vcpu_nr(cpu)];
	}
}

#ifdef CONFIG_XEN_PVHVM
static void __init init_hvm_pv_info(void)
{
	int major, minor;
	uint32_t eax, ebx, ecx, edx, base;

	base = xen_cpuid_base();
	eax = cpuid_eax(base + 1);

	major = eax >> 16;
	minor = eax & 0xffff;
	printk(KERN_INFO "Xen version %d.%d.\n", major, minor);

	xen_domain_type = XEN_HVM_DOMAIN;

	/* PVH set up hypercall page in xen_prepare_pvh(). */
	if (xen_pvh_domain())
		pv_info.name = "Xen PVH";
	else {
		u64 pfn;
		uint32_t msr;

		pv_info.name = "Xen HVM";
		msr = cpuid_ebx(base + 2);
		pfn = __pa(hypercall_page);
		wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));
	}

	xen_setup_features();

	cpuid(base + 4, &eax, &ebx, &ecx, &edx);
	if (eax & XEN_HVM_CPUID_VCPU_ID_PRESENT)
		this_cpu_write(xen_vcpu_id, ebx);
	else
		this_cpu_write(xen_vcpu_id, smp_processor_id());
}
#endif

static int xen_cpu_up_prepare(unsigned int cpu)
{
	int rc;

	if (xen_hvm_domain()) {
		/*
		 * This can happen if CPU was offlined earlier and
		 * offlining timed out in common_cpu_die().
		 */
		if (cpu_report_state(cpu) == CPU_DEAD_FROZEN) {
			xen_smp_intr_free(cpu);
			xen_uninit_lock_cpu(cpu);
		}

		if (cpu_acpi_id(cpu) != U32_MAX)
			per_cpu(xen_vcpu_id, cpu) = cpu_acpi_id(cpu);
		else
			per_cpu(xen_vcpu_id, cpu) = cpu;
		xen_vcpu_setup(cpu);
	}

	if (xen_pv_domain() || xen_feature(XENFEAT_hvm_safe_pvclock))
		xen_setup_timer(cpu);

	rc = xen_smp_intr_init(cpu);
	if (rc) {
		WARN(1, "xen_smp_intr_init() for CPU %d failed: %d\n",
		     cpu, rc);
		return rc;
	}
	return 0;
}

static int xen_cpu_dead(unsigned int cpu)
{
	xen_smp_intr_free(cpu);

	if (xen_pv_domain() || xen_feature(XENFEAT_hvm_safe_pvclock))
		xen_teardown_timer(cpu);

	return 0;
}

static int xen_cpu_up_online(unsigned int cpu)
{
	xen_init_lock_cpu(cpu);
	return 0;
}

#ifdef CONFIG_XEN_PVHVM
#ifdef CONFIG_KEXEC_CORE
static void xen_hvm_shutdown(void)
{
	native_machine_shutdown();
	if (kexec_in_progress)
		xen_reboot(SHUTDOWN_soft_reset);
}

static void xen_hvm_crash_shutdown(struct pt_regs *regs)
{
	native_machine_crash_shutdown(regs);
	xen_reboot(SHUTDOWN_soft_reset);
}
#endif

static void __init xen_hvm_guest_init(void)
{
	if (xen_pv_domain())
		return;

	init_hvm_pv_info();

	xen_hvm_init_shared_info();

	xen_panic_handler_init();

	BUG_ON(!xen_feature(XENFEAT_hvm_callback_vector));

	xen_hvm_smp_init();
	WARN_ON(xen_cpuhp_setup());
	xen_unplug_emulated_devices();
	x86_init.irqs.intr_init = xen_init_IRQ;
	xen_hvm_init_time_ops();
	xen_hvm_init_mmu_ops();

	if (xen_pvh_domain())
		machine_ops.emergency_restart = xen_emergency_restart;
#ifdef CONFIG_KEXEC_CORE
	machine_ops.shutdown = xen_hvm_shutdown;
	machine_ops.crash_shutdown = xen_hvm_crash_shutdown;
#endif
}
#endif

static bool xen_nopv = false;
static __init int xen_parse_nopv(char *arg)
{
       xen_nopv = true;
       return 0;
}
early_param("xen_nopv", xen_parse_nopv);

static uint32_t __init xen_platform(void)
{
	if (xen_nopv)
		return 0;

	return xen_cpuid_base();
}

bool xen_hvm_need_lapic(void)
{
	if (xen_nopv)
		return false;
	if (xen_pv_domain())
		return false;
	if (!xen_hvm_domain())
		return false;
	if (xen_feature(XENFEAT_hvm_pirqs))
		return false;
	return true;
}
EXPORT_SYMBOL_GPL(xen_hvm_need_lapic);

static void xen_set_cpu_features(struct cpuinfo_x86 *c)
{
	if (xen_pv_domain()) {
		clear_cpu_bug(c, X86_BUG_SYSRET_SS_ATTRS);
		set_cpu_cap(c, X86_FEATURE_XENPV);
	}
}

static void xen_pin_vcpu(int cpu)
{
	static bool disable_pinning;
	struct sched_pin_override pin_override;
	int ret;

	if (disable_pinning)
		return;

	pin_override.pcpu = cpu;
	ret = HYPERVISOR_sched_op(SCHEDOP_pin_override, &pin_override);

	/* Ignore errors when removing override. */
	if (cpu < 0)
		return;

	switch (ret) {
	case -ENOSYS:
		pr_warn("Unable to pin on physical cpu %d. In case of problems consider vcpu pinning.\n",
			cpu);
		disable_pinning = true;
		break;
	case -EPERM:
		WARN(1, "Trying to pin vcpu without having privilege to do so\n");
		disable_pinning = true;
		break;
	case -EINVAL:
	case -EBUSY:
		pr_warn("Physical cpu %d not available for pinning. Check Xen cpu configuration.\n",
			cpu);
		break;
	case 0:
		break;
	default:
		WARN(1, "rc %d while trying to pin vcpu\n", ret);
		disable_pinning = true;
	}
}

const struct hypervisor_x86 x86_hyper_xen = {
	.name			= "Xen",
	.detect			= xen_platform,
#ifdef CONFIG_XEN_PVHVM
	.init_platform		= xen_hvm_guest_init,
#endif
	.x2apic_available	= xen_x2apic_para_available,
	.set_cpu_features       = xen_set_cpu_features,
	.pin_vcpu               = xen_pin_vcpu,
};
EXPORT_SYMBOL(x86_hyper_xen);

#ifdef CONFIG_HOTPLUG_CPU
void xen_arch_register_cpu(int num)
{
	arch_register_cpu(num);
}
EXPORT_SYMBOL(xen_arch_register_cpu);

void xen_arch_unregister_cpu(int num)
{
	arch_unregister_cpu(num);
}
EXPORT_SYMBOL(xen_arch_unregister_cpu);
#endif