x86.c

	vcpu->run->io.data_offset = KVM_PIO_PAGE_OFFSET * PAGE_SIZE;
	vcpu->run->io.count = vcpu->arch.pio.count = vcpu->arch.pio.cur_count = 1;
	vcpu->run->io.port = vcpu->arch.pio.port = port;
	vcpu->arch.pio.in = in;
	vcpu->arch.pio.string = 0;
	vcpu->arch.pio.down = 0;
	vcpu->arch.pio.guest_page_offset = 0;
	vcpu->arch.pio.rep = 0;

	kvm_x86_ops->cache_regs(vcpu);
	memcpy(vcpu->arch.pio_data, &vcpu->arch.regs[VCPU_REGS_RAX], 4);
	kvm_x86_ops->decache_regs(vcpu);

	kvm_x86_ops->skip_emulated_instruction(vcpu);

	pio_dev = vcpu_find_pio_dev(vcpu, port);
	if (pio_dev) {
		kernel_pio(pio_dev, vcpu, vcpu->arch.pio_data);
		complete_pio(vcpu);
		return 1;
	}
	return 0;
}
EXPORT_SYMBOL_GPL(kvm_emulate_pio);

int kvm_emulate_pio_string(struct kvm_vcpu *vcpu, struct kvm_run *run, int in,
		  int size, unsigned long count, int down,
		  gva_t address, int rep, unsigned port)
{
	unsigned now, in_page;
	int i, ret = 0;
	int nr_pages = 1;
	struct page *page;
	struct kvm_io_device *pio_dev;

	vcpu->run->exit_reason = KVM_EXIT_IO;
	vcpu->run->io.direction = in ? KVM_EXIT_IO_IN : KVM_EXIT_IO_OUT;
	vcpu->run->io.size = vcpu->arch.pio.size = size;
	vcpu->run->io.data_offset = KVM_PIO_PAGE_OFFSET * PAGE_SIZE;
	vcpu->run->io.count = vcpu->arch.pio.count = vcpu->arch.pio.cur_count = count;
	vcpu->run->io.port = vcpu->arch.pio.port = port;
	vcpu->arch.pio.in = in;
	vcpu->arch.pio.string = 1;
	vcpu->arch.pio.down = down;
	vcpu->arch.pio.guest_page_offset = offset_in_page(address);
	vcpu->arch.pio.rep = rep;

	if (!count) {
		kvm_x86_ops->skip_emulated_instruction(vcpu);
		return 1;
	}

	if (!down)
		in_page = PAGE_SIZE - offset_in_page(address);
	else
		in_page = offset_in_page(address) + size;
	now = min(count, (unsigned long)in_page / size);
	if (!now) {
		/*
		 * String I/O straddles page boundary.  Pin two guest pages
		 * so that we satisfy atomicity constraints.  Do just one
		 * transaction to avoid complexity.
		 */
		nr_pages = 2;
		now = 1;
	}
	if (down) {
		/*
		 * String I/O in reverse.  Yuck.  Kill the guest, fix later.
		 */
		pr_unimpl(vcpu, "guest string pio down\n");
		kvm_inject_gp(vcpu, 0);
		return 1;
	}
	vcpu->run->io.count = now;
	vcpu->arch.pio.cur_count = now;

	if (vcpu->arch.pio.cur_count == vcpu->arch.pio.count)
		kvm_x86_ops->skip_emulated_instruction(vcpu);

	for (i = 0; i < nr_pages; ++i) {
		mutex_lock(&vcpu->kvm->lock);
		page = gva_to_page(vcpu, address + i * PAGE_SIZE);
		vcpu->arch.pio.guest_pages[i] = page;
		mutex_unlock(&vcpu->kvm->lock);
		if (!page) {
			kvm_inject_gp(vcpu, 0);
			free_pio_guest_pages(vcpu);
			return 1;
		}
	}

	pio_dev = vcpu_find_pio_dev(vcpu, port);
	if (!vcpu->arch.pio.in) {
		/* string PIO write */
		ret = pio_copy_data(vcpu);
		if (ret >= 0 && pio_dev) {
			pio_string_write(pio_dev, vcpu);
			complete_pio(vcpu);
			if (vcpu->arch.pio.count == 0)
				ret = 1;
		}
	} else if (pio_dev)
		pr_unimpl(vcpu, "no string pio read support yet, "
		       "port %x size %d count %ld\n",
			port, size, count);

	return ret;
}
EXPORT_SYMBOL_GPL(kvm_emulate_pio_string);

int kvm_arch_init(void *opaque)
{
	int r;
	struct kvm_x86_ops *ops = (struct kvm_x86_ops *)opaque;

	r = kvm_mmu_module_init();
	if (r)
		goto out_fail;

	kvm_init_msr_list();

	if (kvm_x86_ops) {
		printk(KERN_ERR "kvm: already loaded the other module\n");
		r = -EEXIST;
		goto out;
	}

	if (!ops->cpu_has_kvm_support()) {
		printk(KERN_ERR "kvm: no hardware support\n");
		r = -EOPNOTSUPP;
		goto out;
	}
	if (ops->disabled_by_bios()) {
		printk(KERN_ERR "kvm: disabled by bios\n");
		r = -EOPNOTSUPP;
		goto out;
	}

	kvm_x86_ops = ops;
	kvm_mmu_set_nonpresent_ptes(0ull, 0ull);
	return 0;

out:
	kvm_mmu_module_exit();
out_fail:
	return r;
}

void kvm_arch_exit(void)
{
	kvm_x86_ops = NULL;
	kvm_mmu_module_exit();
}

int kvm_emulate_halt(struct kvm_vcpu *vcpu)
{
	++vcpu->stat.halt_exits;
	if (irqchip_in_kernel(vcpu->kvm)) {
		vcpu->arch.mp_state = VCPU_MP_STATE_HALTED;
		kvm_vcpu_block(vcpu);
		if (vcpu->arch.mp_state != VCPU_MP_STATE_RUNNABLE)
			return -EINTR;
		return 1;
	} else {
		vcpu->run->exit_reason = KVM_EXIT_HLT;
		return 0;
	}
}
EXPORT_SYMBOL_GPL(kvm_emulate_halt);

int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
{
	unsigned long nr, a0, a1, a2, a3, ret;

	kvm_x86_ops->cache_regs(vcpu);

	nr = vcpu->arch.regs[VCPU_REGS_RAX];
	a0 = vcpu->arch.regs[VCPU_REGS_RBX];
	a1 = vcpu->arch.regs[VCPU_REGS_RCX];
	a2 = vcpu->arch.regs[VCPU_REGS_RDX];
	a3 = vcpu->arch.regs[VCPU_REGS_RSI];

	if (!is_long_mode(vcpu)) {
		nr &= 0xFFFFFFFF;
		a0 &= 0xFFFFFFFF;
		a1 &= 0xFFFFFFFF;
		a2 &= 0xFFFFFFFF;
		a3 &= 0xFFFFFFFF;
	}

	switch (nr) {
	default:
		ret = -KVM_ENOSYS;
		break;
	}
	vcpu->arch.regs[VCPU_REGS_RAX] = ret;
	kvm_x86_ops->decache_regs(vcpu);
	return 0;
}
EXPORT_SYMBOL_GPL(kvm_emulate_hypercall);

int kvm_fix_hypercall(struct kvm_vcpu *vcpu)
{
	char instruction[3];
	int ret = 0;

	mutex_lock(&vcpu->kvm->lock);

	/*
	 * Blow out the MMU to ensure that no other VCPU has an active mapping
	 * to ensure that the updated hypercall appears atomically across all
	 * VCPUs.
	 */
	kvm_mmu_zap_all(vcpu->kvm);

	kvm_x86_ops->cache_regs(vcpu);
	kvm_x86_ops->patch_hypercall(vcpu, instruction);
	if (emulator_write_emulated(vcpu->arch.rip, instruction, 3, vcpu)
	    != X86EMUL_CONTINUE)
		ret = -EFAULT;

	mutex_unlock(&vcpu->kvm->lock);

	return ret;
}

static u64 mk_cr_64(u64 curr_cr, u32 new_val)
{
	return (curr_cr & ~((1ULL << 32) - 1)) | new_val;
}

void realmode_lgdt(struct kvm_vcpu *vcpu, u16 limit, unsigned long base)
{
	struct descriptor_table dt = { limit, base };

	kvm_x86_ops->set_gdt(vcpu, &dt);
}

void realmode_lidt(struct kvm_vcpu *vcpu, u16 limit, unsigned long base)
{
	struct descriptor_table dt = { limit, base };

	kvm_x86_ops->set_idt(vcpu, &dt);
}

void realmode_lmsw(struct kvm_vcpu *vcpu, unsigned long msw,
		   unsigned long *rflags)
{
	lmsw(vcpu, msw);
	*rflags = kvm_x86_ops->get_rflags(vcpu);
}

unsigned long realmode_get_cr(struct kvm_vcpu *vcpu, int cr)
{
	kvm_x86_ops->decache_cr4_guest_bits(vcpu);
	switch (cr) {
	case 0:
		return vcpu->arch.cr0;
	case 2:
		return vcpu->arch.cr2;
	case 3:
		return vcpu->arch.cr3;
	case 4:
		return vcpu->arch.cr4;
	case 8:
		return get_cr8(vcpu);
	default:
		vcpu_printf(vcpu, "%s: unexpected cr %u\n", __FUNCTION__, cr);
		return 0;
	}
}

void realmode_set_cr(struct kvm_vcpu *vcpu, int cr, unsigned long val,
		     unsigned long *rflags)
{
	switch (cr) {
	case 0:
		set_cr0(vcpu, mk_cr_64(vcpu->arch.cr0, val));
		*rflags = kvm_x86_ops->get_rflags(vcpu);
		break;
	case 2:
		vcpu->arch.cr2 = val;
		break;
	case 3:
		set_cr3(vcpu, val);
		break;
	case 4:
		set_cr4(vcpu, mk_cr_64(vcpu->arch.cr4, val));
		break;
	case 8:
		set_cr8(vcpu, val & 0xfUL);
		break;
	default:
		vcpu_printf(vcpu, "%s: unexpected cr %u\n", __FUNCTION__, cr);
	}
}

static int move_to_next_stateful_cpuid_entry(struct kvm_vcpu *vcpu, int i)
{
	struct kvm_cpuid_entry2 *e = &vcpu->arch.cpuid_entries[i];
	int j, nent = vcpu->arch.cpuid_nent;

	e->flags &= ~KVM_CPUID_FLAG_STATE_READ_NEXT;
	/* when no next entry is found, the current entry[i] is reselected */
	for (j = i + 1; j == i; j = (j + 1) % nent) {
		struct kvm_cpuid_entry2 *ej = &vcpu->arch.cpuid_entries[j];
		if (ej->function == e->function) {
			ej->flags |= KVM_CPUID_FLAG_STATE_READ_NEXT;
			return j;
		}
	}
	return 0; /* silence gcc, even though control never reaches here */
}

/* find an entry with matching function, matching index (if needed), and that
 * should be read next (if it's stateful) */
static int is_matching_cpuid_entry(struct kvm_cpuid_entry2 *e,
	u32 function, u32 index)
{
	if (e->function != function)
		return 0;
	if ((e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX) && e->index != index)
		return 0;
	if ((e->flags & KVM_CPUID_FLAG_STATEFUL_FUNC) &&
		!(e->flags & KVM_CPUID_FLAG_STATE_READ_NEXT))
		return 0;
	return 1;
}

void kvm_emulate_cpuid(struct kvm_vcpu *vcpu)
{
	int i;
	u32 function, index;
	struct kvm_cpuid_entry2 *e, *best;

	kvm_x86_ops->cache_regs(vcpu);
	function = vcpu->arch.regs[VCPU_REGS_RAX];
	index = vcpu->arch.regs[VCPU_REGS_RCX];
	vcpu->arch.regs[VCPU_REGS_RAX] = 0;
	vcpu->arch.regs[VCPU_REGS_RBX] = 0;
	vcpu->arch.regs[VCPU_REGS_RCX] = 0;
	vcpu->arch.regs[VCPU_REGS_RDX] = 0;
	best = NULL;
	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
		e = &vcpu->arch.cpuid_entries[i];
		if (is_matching_cpuid_entry(e, function, index)) {
			if (e->flags & KVM_CPUID_FLAG_STATEFUL_FUNC)
				move_to_next_stateful_cpuid_entry(vcpu, i);
			best = e;
			break;
		}
		/*
		 * Both basic or both extended?
		 */
		if (((e->function ^ function) & 0x80000000) == 0)
			if (!best || e->function > best->function)
				best = e;
	}
	if (best) {
		vcpu->arch.regs[VCPU_REGS_RAX] = best->eax;
		vcpu->arch.regs[VCPU_REGS_RBX] = best->ebx;
		vcpu->arch.regs[VCPU_REGS_RCX] = best->ecx;
		vcpu->arch.regs[VCPU_REGS_RDX] = best->edx;
	}
	kvm_x86_ops->decache_regs(vcpu);
	kvm_x86_ops->skip_emulated_instruction(vcpu);
}
EXPORT_SYMBOL_GPL(kvm_emulate_cpuid);

/*
 * Check if userspace requested an interrupt window, and that the
 * interrupt window is open.
 *
 * No need to exit to userspace if we already have an interrupt queued.
 */
static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
					  struct kvm_run *kvm_run)
{
	return (!vcpu->arch.irq_summary &&
		kvm_run->request_interrupt_window &&
		vcpu->arch.interrupt_window_open &&
		(kvm_x86_ops->get_rflags(vcpu) & X86_EFLAGS_IF));
}

static void post_kvm_run_save(struct kvm_vcpu *vcpu,
			      struct kvm_run *kvm_run)
{
	kvm_run->if_flag = (kvm_x86_ops->get_rflags(vcpu) & X86_EFLAGS_IF) != 0;
	kvm_run->cr8 = get_cr8(vcpu);
	kvm_run->apic_base = kvm_get_apic_base(vcpu);
	if (irqchip_in_kernel(vcpu->kvm))
		kvm_run->ready_for_interrupt_injection = 1;
	else
		kvm_run->ready_for_interrupt_injection =
					(vcpu->arch.interrupt_window_open &&
					 vcpu->arch.irq_summary == 0);
}

static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
{
	int r;

	if (unlikely(vcpu->arch.mp_state == VCPU_MP_STATE_SIPI_RECEIVED)) {
		pr_debug("vcpu %d received sipi with vector # %x\n",
		       vcpu->vcpu_id, vcpu->arch.sipi_vector);
		kvm_lapic_reset(vcpu);
		r = kvm_x86_ops->vcpu_reset(vcpu);
		if (r)
			return r;
		vcpu->arch.mp_state = VCPU_MP_STATE_RUNNABLE;
	}

preempted:
	if (vcpu->guest_debug.enabled)
		kvm_x86_ops->guest_debug_pre(vcpu);

again:
	r = kvm_mmu_reload(vcpu);
	if (unlikely(r))
		goto out;

	kvm_inject_pending_timer_irqs(vcpu);

	preempt_disable();

	kvm_x86_ops->prepare_guest_switch(vcpu);
	kvm_load_guest_fpu(vcpu);

	local_irq_disable();

	if (signal_pending(current)) {
		local_irq_enable();
		preempt_enable();
		r = -EINTR;
		kvm_run->exit_reason = KVM_EXIT_INTR;
		++vcpu->stat.signal_exits;
		goto out;
	}

	if (vcpu->arch.exception.pending)
		__queue_exception(vcpu);
	else if (irqchip_in_kernel(vcpu->kvm))
		kvm_x86_ops->inject_pending_irq(vcpu);
	else
		kvm_x86_ops->inject_pending_vectors(vcpu, kvm_run);

	vcpu->guest_mode = 1;
	kvm_guest_enter();

	if (vcpu->requests)
		if (test_and_clear_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests))
			kvm_x86_ops->tlb_flush(vcpu);

	kvm_x86_ops->run(vcpu, kvm_run);

	vcpu->guest_mode = 0;
	local_irq_enable();

	++vcpu->stat.exits;

	/*
	 * We must have an instruction between local_irq_enable() and
	 * kvm_guest_exit(), so the timer interrupt isn't delayed by
	 * the interrupt shadow.  The stat.exits increment will do nicely.
	 * But we need to prevent reordering, hence this barrier():
	 */
	barrier();

	kvm_guest_exit();

	preempt_enable();

	/*
	 * Profile KVM exit RIPs:
	 */
	if (unlikely(prof_on == KVM_PROFILING)) {
		kvm_x86_ops->cache_regs(vcpu);
		profile_hit(KVM_PROFILING, (void *)vcpu->arch.rip);
	}

	if (vcpu->arch.exception.pending && kvm_x86_ops->exception_injected(vcpu))
		vcpu->arch.exception.pending = false;

	r = kvm_x86_ops->handle_exit(kvm_run, vcpu);

	if (r > 0) {
		if (dm_request_for_irq_injection(vcpu, kvm_run)) {
			r = -EINTR;
			kvm_run->exit_reason = KVM_EXIT_INTR;
			++vcpu->stat.request_irq_exits;
			goto out;
		}
		if (!need_resched())
			goto again;
	}

out:
	if (r > 0) {
		kvm_resched(vcpu);
		goto preempted;
	}

	post_kvm_run_save(vcpu, kvm_run);

	return r;
}

int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
{
	int r;
	sigset_t sigsaved;

	vcpu_load(vcpu);

	if (unlikely(vcpu->arch.mp_state == VCPU_MP_STATE_UNINITIALIZED)) {
		kvm_vcpu_block(vcpu);
		vcpu_put(vcpu);
		return -EAGAIN;
	}

	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);

	/* re-sync apic's tpr */
	if (!irqchip_in_kernel(vcpu->kvm))
		set_cr8(vcpu, kvm_run->cr8);

	if (vcpu->arch.pio.cur_count) {
		r = complete_pio(vcpu);
		if (r)
			goto out;
	}
#if CONFIG_HAS_IOMEM
	if (vcpu->mmio_needed) {
		memcpy(vcpu->mmio_data, kvm_run->mmio.data, 8);
		vcpu->mmio_read_completed = 1;
		vcpu->mmio_needed = 0;
		r = emulate_instruction(vcpu, kvm_run,
					vcpu->arch.mmio_fault_cr2, 0, 1);
		if (r == EMULATE_DO_MMIO) {
			/*
			 * Read-modify-write.  Back to userspace.
			 */
			r = 0;
			goto out;
		}
	}
#endif
	if (kvm_run->exit_reason == KVM_EXIT_HYPERCALL) {
		kvm_x86_ops->cache_regs(vcpu);
		vcpu->arch.regs[VCPU_REGS_RAX] = kvm_run->hypercall.ret;
		kvm_x86_ops->decache_regs(vcpu);
	}

	r = __vcpu_run(vcpu, kvm_run);

out:
	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &sigsaved, NULL);

	vcpu_put(vcpu);
	return r;
}

int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	vcpu_load(vcpu);

	kvm_x86_ops->cache_regs(vcpu);

	regs->rax = vcpu->arch.regs[VCPU_REGS_RAX];
	regs->rbx = vcpu->arch.regs[VCPU_REGS_RBX];
	regs->rcx = vcpu->arch.regs[VCPU_REGS_RCX];
	regs->rdx = vcpu->arch.regs[VCPU_REGS_RDX];
	regs->rsi = vcpu->arch.regs[VCPU_REGS_RSI];
	regs->rdi = vcpu->arch.regs[VCPU_REGS_RDI];
	regs->rsp = vcpu->arch.regs[VCPU_REGS_RSP];
	regs->rbp = vcpu->arch.regs[VCPU_REGS_RBP];
#ifdef CONFIG_X86_64
	regs->r8 = vcpu->arch.regs[VCPU_REGS_R8];
	regs->r9 = vcpu->arch.regs[VCPU_REGS_R9];
	regs->r10 = vcpu->arch.regs[VCPU_REGS_R10];
	regs->r11 = vcpu->arch.regs[VCPU_REGS_R11];
	regs->r12 = vcpu->arch.regs[VCPU_REGS_R12];
	regs->r13 = vcpu->arch.regs[VCPU_REGS_R13];
	regs->r14 = vcpu->arch.regs[VCPU_REGS_R14];
	regs->r15 = vcpu->arch.regs[VCPU_REGS_R15];
#endif

	regs->rip = vcpu->arch.rip;
	regs->rflags = kvm_x86_ops->get_rflags(vcpu);

	/*
	 * Don't leak debug flags in case they were set for guest debugging
	 */
	if (vcpu->guest_debug.enabled && vcpu->guest_debug.singlestep)
		regs->rflags &= ~(X86_EFLAGS_TF | X86_EFLAGS_RF);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	vcpu_load(vcpu);

	vcpu->arch.regs[VCPU_REGS_RAX] = regs->rax;
	vcpu->arch.regs[VCPU_REGS_RBX] = regs->rbx;
	vcpu->arch.regs[VCPU_REGS_RCX] = regs->rcx;
	vcpu->arch.regs[VCPU_REGS_RDX] = regs->rdx;
	vcpu->arch.regs[VCPU_REGS_RSI] = regs->rsi;
	vcpu->arch.regs[VCPU_REGS_RDI] = regs->rdi;
	vcpu->arch.regs[VCPU_REGS_RSP] = regs->rsp;
	vcpu->arch.regs[VCPU_REGS_RBP] = regs->rbp;
#ifdef CONFIG_X86_64
	vcpu->arch.regs[VCPU_REGS_R8] = regs->r8;
	vcpu->arch.regs[VCPU_REGS_R9] = regs->r9;
	vcpu->arch.regs[VCPU_REGS_R10] = regs->r10;
	vcpu->arch.regs[VCPU_REGS_R11] = regs->r11;
	vcpu->arch.regs[VCPU_REGS_R12] = regs->r12;
	vcpu->arch.regs[VCPU_REGS_R13] = regs->r13;
	vcpu->arch.regs[VCPU_REGS_R14] = regs->r14;
	vcpu->arch.regs[VCPU_REGS_R15] = regs->r15;
#endif

	vcpu->arch.rip = regs->rip;
	kvm_x86_ops->set_rflags(vcpu, regs->rflags);

	kvm_x86_ops->decache_regs(vcpu);

	vcpu_put(vcpu);

	return 0;
}

static void get_segment(struct kvm_vcpu *vcpu,
			struct kvm_segment *var, int seg)
{
	return kvm_x86_ops->get_segment(vcpu, var, seg);
}

void kvm_get_cs_db_l_bits(struct kvm_vcpu *vcpu, int *db, int *l)
{
	struct kvm_segment cs;

	get_segment(vcpu, &cs, VCPU_SREG_CS);
	*db = cs.db;
	*l = cs.l;
}
EXPORT_SYMBOL_GPL(kvm_get_cs_db_l_bits);

int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
	struct descriptor_table dt;
	int pending_vec;

	vcpu_load(vcpu);

	get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
	get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
	get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
	get_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
	get_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
	get_segment(vcpu, &sregs->ss, VCPU_SREG_SS);

	get_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
	get_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);

	kvm_x86_ops->get_idt(vcpu, &dt);
	sregs->idt.limit = dt.limit;
	sregs->idt.base = dt.base;
	kvm_x86_ops->get_gdt(vcpu, &dt);
	sregs->gdt.limit = dt.limit;
	sregs->gdt.base = dt.base;

	kvm_x86_ops->decache_cr4_guest_bits(vcpu);
	sregs->cr0 = vcpu->arch.cr0;
	sregs->cr2 = vcpu->arch.cr2;
	sregs->cr3 = vcpu->arch.cr3;
	sregs->cr4 = vcpu->arch.cr4;
	sregs->cr8 = get_cr8(vcpu);
	sregs->efer = vcpu->arch.shadow_efer;
	sregs->apic_base = kvm_get_apic_base(vcpu);

	if (irqchip_in_kernel(vcpu->kvm)) {
		memset(sregs->interrupt_bitmap, 0,
		       sizeof sregs->interrupt_bitmap);
		pending_vec = kvm_x86_ops->get_irq(vcpu);
		if (pending_vec >= 0)
			set_bit(pending_vec,
				(unsigned long *)sregs->interrupt_bitmap);
	} else
		memcpy(sregs->interrupt_bitmap, vcpu->arch.irq_pending,
		       sizeof sregs->interrupt_bitmap);

	vcpu_put(vcpu);

	return 0;
}

static void set_segment(struct kvm_vcpu *vcpu,
			struct kvm_segment *var, int seg)
{
	return kvm_x86_ops->set_segment(vcpu, var, seg);
}

int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
	int mmu_reset_needed = 0;
	int i, pending_vec, max_bits;
	struct descriptor_table dt;

	vcpu_load(vcpu);

	dt.limit = sregs->idt.limit;
	dt.base = sregs->idt.base;
	kvm_x86_ops->set_idt(vcpu, &dt);
	dt.limit = sregs->gdt.limit;
	dt.base = sregs->gdt.base;
	kvm_x86_ops->set_gdt(vcpu, &dt);

	vcpu->arch.cr2 = sregs->cr2;
	mmu_reset_needed |= vcpu->arch.cr3 != sregs->cr3;
	vcpu->arch.cr3 = sregs->cr3;

	set_cr8(vcpu, sregs->cr8);

	mmu_reset_needed |= vcpu->arch.shadow_efer != sregs->efer;
#ifdef CONFIG_X86_64
	kvm_x86_ops->set_efer(vcpu, sregs->efer);
#endif
	kvm_set_apic_base(vcpu, sregs->apic_base);

	kvm_x86_ops->decache_cr4_guest_bits(vcpu);

	mmu_reset_needed |= vcpu->arch.cr0 != sregs->cr0;
	vcpu->arch.cr0 = sregs->cr0;
	kvm_x86_ops->set_cr0(vcpu, sregs->cr0);

	mmu_reset_needed |= vcpu->arch.cr4 != sregs->cr4;
	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
	if (!is_long_mode(vcpu) && is_pae(vcpu))
		load_pdptrs(vcpu, vcpu->arch.cr3);

	if (mmu_reset_needed)
		kvm_mmu_reset_context(vcpu);

	if (!irqchip_in_kernel(vcpu->kvm)) {
		memcpy(vcpu->arch.irq_pending, sregs->interrupt_bitmap,
		       sizeof vcpu->arch.irq_pending);
		vcpu->arch.irq_summary = 0;
		for (i = 0; i < ARRAY_SIZE(vcpu->arch.irq_pending); ++i)
			if (vcpu->arch.irq_pending[i])
				__set_bit(i, &vcpu->arch.irq_summary);
	} else {
		max_bits = (sizeof sregs->interrupt_bitmap) << 3;
		pending_vec = find_first_bit(
			(const unsigned long *)sregs->interrupt_bitmap,
			max_bits);
		/* Only pending external irq is handled here */
		if (pending_vec < max_bits) {
			kvm_x86_ops->set_irq(vcpu, pending_vec);
			pr_debug("Set back pending irq %d\n",
				 pending_vec);
		}
	}

	set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
	set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
	set_segment(vcpu, &sregs->es, VCPU_SREG_ES);
	set_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
	set_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
	set_segment(vcpu, &sregs->ss, VCPU_SREG_SS);

	set_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
	set_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
				    struct kvm_debug_guest *dbg)
{
	int r;

	vcpu_load(vcpu);

	r = kvm_x86_ops->set_guest_debug(vcpu, dbg);

	vcpu_put(vcpu);

	return r;
}

/*
 * fxsave fpu state.  Taken from x86_64/processor.h.  To be killed when
 * we have asm/x86/processor.h
 */
struct fxsave {
	u16	cwd;
	u16	swd;
	u16	twd;
	u16	fop;
	u64	rip;
	u64	rdp;
	u32	mxcsr;
	u32	mxcsr_mask;
	u32	st_space[32];	/* 8*16 bytes for each FP-reg = 128 bytes */
#ifdef CONFIG_X86_64
	u32	xmm_space[64];	/* 16*16 bytes for each XMM-reg = 256 bytes */
#else
	u32	xmm_space[32];	/* 8*16 bytes for each XMM-reg = 128 bytes */
#endif
};

/*
 * Translate a guest virtual address to a guest physical address.
 */
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr)
{
	unsigned long vaddr = tr->linear_address;
	gpa_t gpa;

	vcpu_load(vcpu);
	mutex_lock(&vcpu->kvm->lock);
	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, vaddr);
	tr->physical_address = gpa;
	tr->valid = gpa != UNMAPPED_GVA;
	tr->writeable = 1;
	tr->usermode = 0;
	mutex_unlock(&vcpu->kvm->lock);
	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	struct fxsave *fxsave = (struct fxsave *)&vcpu->arch.guest_fx_image;

	vcpu_load(vcpu);

	memcpy(fpu->fpr, fxsave->st_space, 128);
	fpu->fcw = fxsave->cwd;
	fpu->fsw = fxsave->swd;
	fpu->ftwx = fxsave->twd;
	fpu->last_opcode = fxsave->fop;
	fpu->last_ip = fxsave->rip;
	fpu->last_dp = fxsave->rdp;
	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	struct fxsave *fxsave = (struct fxsave *)&vcpu->arch.guest_fx_image;

	vcpu_load(vcpu);

	memcpy(fxsave->st_space, fpu->fpr, 128);
	fxsave->cwd = fpu->fcw;
	fxsave->swd = fpu->fsw;
	fxsave->twd = fpu->ftwx;
	fxsave->fop = fpu->last_opcode;
	fxsave->rip = fpu->last_ip;
	fxsave->rdp = fpu->last_dp;
	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);

	vcpu_put(vcpu);

	return 0;
}

void fx_init(struct kvm_vcpu *vcpu)
{
	unsigned after_mxcsr_mask;

	/* Initialize guest FPU by resetting ours and saving into guest's */
	preempt_disable();
	fx_save(&vcpu->arch.host_fx_image);
	fpu_init();
	fx_save(&vcpu->arch.guest_fx_image);
	fx_restore(&vcpu->arch.host_fx_image);
	preempt_enable();

	vcpu->arch.cr0 |= X86_CR0_ET;
	after_mxcsr_mask = offsetof(struct i387_fxsave_struct, st_space);
	vcpu->arch.guest_fx_image.mxcsr = 0x1f80;
	memset((void *)&vcpu->arch.guest_fx_image + after_mxcsr_mask,
	       0, sizeof(struct i387_fxsave_struct) - after_mxcsr_mask);
}
EXPORT_SYMBOL_GPL(fx_init);

void kvm_load_guest_fpu(struct kvm_vcpu *vcpu)
{
	if (!vcpu->fpu_active || vcpu->guest_fpu_loaded)
		return;

	vcpu->guest_fpu_loaded = 1;
	fx_save(&vcpu->arch.host_fx_image);
	fx_restore(&vcpu->arch.guest_fx_image);
}
EXPORT_SYMBOL_GPL(kvm_load_guest_fpu);

void kvm_put_guest_fpu(struct kvm_vcpu *vcpu)
{
	if (!vcpu->guest_fpu_loaded)
		return;

	vcpu->guest_fpu_loaded = 0;
	fx_save(&vcpu->arch.guest_fx_image);
	fx_restore(&vcpu->arch.host_fx_image);
	++vcpu->stat.fpu_reload;
}
EXPORT_SYMBOL_GPL(kvm_put_guest_fpu);

void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
	kvm_x86_ops->vcpu_free(vcpu);
}

struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
						unsigned int id)
{
	return kvm_x86_ops->vcpu_create(kvm, id);
}

int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
{
	int r;

	/* We do fxsave: this must be aligned. */
	BUG_ON((unsigned long)&vcpu->arch.host_fx_image & 0xF);

	vcpu_load(vcpu);
	r = kvm_arch_vcpu_reset(vcpu);
	if (r == 0)
		r = kvm_mmu_setup(vcpu);
	vcpu_put(vcpu);
	if (r < 0)
		goto free_vcpu;

	return 0;
free_vcpu:
	kvm_x86_ops->vcpu_free(vcpu);
	return r;
}

void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
{
	vcpu_load(vcpu);
	kvm_mmu_unload(vcpu);
	vcpu_put(vcpu);

	kvm_x86_ops->vcpu_free(vcpu);
}

int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
{
	return kvm_x86_ops->vcpu_reset(vcpu);
}

void kvm_arch_hardware_enable(void *garbage)
{
	kvm_x86_ops->hardware_enable(garbage);
}

void kvm_arch_hardware_disable(void *garbage)
{
	kvm_x86_ops->hardware_disable(garbage);
}

int kvm_arch_hardware_setup(void)
{
	return kvm_x86_ops->hardware_setup();
}

void kvm_arch_hardware_unsetup(void)
{
	kvm_x86_ops->hardware_unsetup();
}

void kvm_arch_check_processor_compat(void *rtn)
{
	kvm_x86_ops->check_processor_compatibility(rtn);
}

int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
	struct page *page;