Merge branch 'linus' into x86/urgent

15 files changed, 691 insertions, 386 deletions

diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c
index 87edf1ceb1d..d02def06ca9 100644
--- a/arch/x86/kernel/kvmclock.c
+++ b/arch/x86/kernel/kvmclock.c
@@ -113,7 +113,7 @@ static void kvm_setup_secondary_clock(void)
 #endif
 
 #ifdef CONFIG_SMP
-void __init kvm_smp_prepare_boot_cpu(void)
+static void __init kvm_smp_prepare_boot_cpu(void)
 {
 	WARN_ON(kvm_register_clock("primary cpu clock"));
 	native_smp_prepare_boot_cpu();
diff --git a/arch/x86/kvm/Makefile b/arch/x86/kvm/Makefile
index c97d35c218d..d0e940bb6f4 100644
--- a/arch/x86/kvm/Makefile
+++ b/arch/x86/kvm/Makefile
@@ -2,7 +2,8 @@
 # Makefile for Kernel-based Virtual Machine module
 #
 
-common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o ioapic.o)
+common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o ioapic.o \
+                coalesced_mmio.o)
 ifeq ($(CONFIG_KVM_TRACE),y)
 common-objs += $(addprefix ../../../virt/kvm/, kvm_trace.o)
 endif
diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c
index 3829aa7b663..c0f7872a912 100644
--- a/arch/x86/kvm/i8254.c
+++ b/arch/x86/kvm/i8254.c
@@ -91,7 +91,7 @@ static void pit_set_gate(struct kvm *kvm, int channel, u32 val)
 	c->gate = val;
 }
 
-int pit_get_gate(struct kvm *kvm, int channel)
+static int pit_get_gate(struct kvm *kvm, int channel)
 {
 	WARN_ON(!mutex_is_locked(&kvm->arch.vpit->pit_state.lock));
 
@@ -193,19 +193,16 @@ static void pit_latch_status(struct kvm *kvm, int channel)
 	}
 }
 
-int __pit_timer_fn(struct kvm_kpit_state *ps)
+static int __pit_timer_fn(struct kvm_kpit_state *ps)
 {
 	struct kvm_vcpu *vcpu0 = ps->pit->kvm->vcpus[0];
 	struct kvm_kpit_timer *pt = &ps->pit_timer;
 
-	atomic_inc(&pt->pending);
-	smp_mb__after_atomic_inc();
-	if (vcpu0) {
+	if (!atomic_inc_and_test(&pt->pending))
 		set_bit(KVM_REQ_PENDING_TIMER, &vcpu0->requests);
-		if (waitqueue_active(&vcpu0->wq)) {
-			vcpu0->arch.mp_state = KVM_MP_STATE_RUNNABLE;
-			wake_up_interruptible(&vcpu0->wq);
-		}
+	if (vcpu0 && waitqueue_active(&vcpu0->wq)) {
+		vcpu0->arch.mp_state = KVM_MP_STATE_RUNNABLE;
+		wake_up_interruptible(&vcpu0->wq);
 	}
 
 	pt->timer.expires = ktime_add_ns(pt->timer.expires, pt->period);
@@ -308,6 +305,7 @@ static void pit_load_count(struct kvm *kvm, int channel, u32 val)
 		create_pit_timer(&ps->pit_timer, val, 0);
 		break;
 	case 2:
+	case 3:
 		create_pit_timer(&ps->pit_timer, val, 1);
 		break;
 	default:
@@ -459,7 +457,8 @@ static void pit_ioport_read(struct kvm_io_device *this,
 	mutex_unlock(&pit_state->lock);
 }
 
-static int pit_in_range(struct kvm_io_device *this, gpa_t addr)
+static int pit_in_range(struct kvm_io_device *this, gpa_t addr,
+			int len, int is_write)
 {
 	return ((addr >= KVM_PIT_BASE_ADDRESS) &&
 		(addr < KVM_PIT_BASE_ADDRESS + KVM_PIT_MEM_LENGTH));
@@ -500,7 +499,8 @@ static void speaker_ioport_read(struct kvm_io_device *this,
 	mutex_unlock(&pit_state->lock);
 }
 
-static int speaker_in_range(struct kvm_io_device *this, gpa_t addr)
+static int speaker_in_range(struct kvm_io_device *this, gpa_t addr,
+			    int len, int is_write)
 {
 	return (addr == KVM_SPEAKER_BASE_ADDRESS);
 }
@@ -575,7 +575,7 @@ void kvm_free_pit(struct kvm *kvm)
 	}
 }
 
-void __inject_pit_timer_intr(struct kvm *kvm)
+static void __inject_pit_timer_intr(struct kvm *kvm)
 {
 	mutex_lock(&kvm->lock);
 	kvm_ioapic_set_irq(kvm->arch.vioapic, 0, 1);
diff --git a/arch/x86/kvm/i8259.c b/arch/x86/kvm/i8259.c
index ab29cf2def4..c31164e8aa4 100644
--- a/arch/x86/kvm/i8259.c
+++ b/arch/x86/kvm/i8259.c
@@ -130,8 +130,10 @@ void kvm_pic_set_irq(void *opaque, int irq, int level)
 {
 	struct kvm_pic *s = opaque;
 
-	pic_set_irq1(&s->pics[irq >> 3], irq & 7, level);
-	pic_update_irq(s);
+	if (irq >= 0 && irq < PIC_NUM_PINS) {
+		pic_set_irq1(&s->pics[irq >> 3], irq & 7, level);
+		pic_update_irq(s);
+	}
 }
 
 /*
@@ -346,7 +348,8 @@ static u32 elcr_ioport_read(void *opaque, u32 addr1)
 	return s->elcr;
 }
 
-static int picdev_in_range(struct kvm_io_device *this, gpa_t addr)
+static int picdev_in_range(struct kvm_io_device *this, gpa_t addr,
+			   int len, int is_write)
 {
 	switch (addr) {
 	case 0x20:
diff --git a/arch/x86/kvm/irq.h b/arch/x86/kvm/irq.h
index 2a15be2275c..7ca47cbb48b 100644
--- a/arch/x86/kvm/irq.h
+++ b/arch/x86/kvm/irq.h
@@ -30,6 +30,8 @@
 #include "ioapic.h"
 #include "lapic.h"
 
+#define PIC_NUM_PINS 16
+
 struct kvm;
 struct kvm_vcpu;
 
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index ebc03f5ae16..73f43de69f6 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -356,8 +356,9 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 	case APIC_DM_SMI:
 		printk(KERN_DEBUG "Ignoring guest SMI\n");
 		break;
+
 	case APIC_DM_NMI:
-		printk(KERN_DEBUG "Ignoring guest NMI\n");
+		kvm_inject_nmi(vcpu);
 		break;
 
 	case APIC_DM_INIT:
@@ -572,6 +573,8 @@ static u32 __apic_read(struct kvm_lapic *apic, unsigned int offset)
 {
 	u32 val = 0;
 
+	KVMTRACE_1D(APIC_ACCESS, apic->vcpu, (u32)offset, handler);
+
 	if (offset >= LAPIC_MMIO_LENGTH)
 		return 0;
 
@@ -695,6 +698,8 @@ static void apic_mmio_write(struct kvm_io_device *this,
 
 	offset &= 0xff0;
 
+	KVMTRACE_1D(APIC_ACCESS, apic->vcpu, (u32)offset, handler);
+
 	switch (offset) {
 	case APIC_ID:		/* Local APIC ID */
 		apic_set_reg(apic, APIC_ID, val);
@@ -780,7 +785,8 @@ static void apic_mmio_write(struct kvm_io_device *this,
 
 }
 
-static int apic_mmio_range(struct kvm_io_device *this, gpa_t addr)
+static int apic_mmio_range(struct kvm_io_device *this, gpa_t addr,
+			   int len, int size)
 {
 	struct kvm_lapic *apic = (struct kvm_lapic *)this->private;
 	int ret = 0;
@@ -939,8 +945,8 @@ static int __apic_timer_fn(struct kvm_lapic *apic)
 	int result = 0;
 	wait_queue_head_t *q = &apic->vcpu->wq;
 
-	atomic_inc(&apic->timer.pending);
-	set_bit(KVM_REQ_PENDING_TIMER, &apic->vcpu->requests);
+	if(!atomic_inc_and_test(&apic->timer.pending))
+		set_bit(KVM_REQ_PENDING_TIMER, &apic->vcpu->requests);
 	if (waitqueue_active(q)) {
 		apic->vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
 		wake_up_interruptible(q);
diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h
index 676c396c9ce..81858881287 100644
--- a/arch/x86/kvm/lapic.h
+++ b/arch/x86/kvm/lapic.h
@@ -31,6 +31,7 @@ void kvm_lapic_reset(struct kvm_vcpu *vcpu);
 u64 kvm_lapic_get_cr8(struct kvm_vcpu *vcpu);
 void kvm_lapic_set_tpr(struct kvm_vcpu *vcpu, unsigned long cr8);
 void kvm_lapic_set_base(struct kvm_vcpu *vcpu, u64 value);
+u64 kvm_lapic_get_base(struct kvm_vcpu *vcpu);
 
 int kvm_apic_match_physical_addr(struct kvm_lapic *apic, u16 dest);
 int kvm_apic_match_logical_addr(struct kvm_lapic *apic, u8 mda);
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 7e7c3969f7a..b0e4ddca6c1 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -66,7 +66,8 @@ static void kvm_mmu_audit(struct kvm_vcpu *vcpu, const char *msg) {}
 #endif
 
 #if defined(MMU_DEBUG) || defined(AUDIT)
-static int dbg = 1;
+static int dbg = 0;
+module_param(dbg, bool, 0644);
 #endif
 
 #ifndef MMU_DEBUG
@@ -776,6 +777,15 @@ static void mmu_page_remove_parent_pte(struct kvm_mmu_page *sp,
 	BUG();
 }
 
+static void nonpaging_prefetch_page(struct kvm_vcpu *vcpu,
+				    struct kvm_mmu_page *sp)
+{
+	int i;
+
+	for (i = 0; i < PT64_ENT_PER_PAGE; ++i)
+		sp->spt[i] = shadow_trap_nonpresent_pte;
+}
+
 static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm *kvm, gfn_t gfn)
 {
 	unsigned index;
@@ -841,7 +851,10 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu,
 	hlist_add_head(&sp->hash_link, bucket);
 	if (!metaphysical)
 		rmap_write_protect(vcpu->kvm, gfn);
-	vcpu->arch.mmu.prefetch_page(vcpu, sp);
+	if (shadow_trap_nonpresent_pte != shadow_notrap_nonpresent_pte)
+		vcpu->arch.mmu.prefetch_page(vcpu, sp);
+	else
+		nonpaging_prefetch_page(vcpu, sp);
 	return sp;
 }
 
@@ -917,14 +930,17 @@ static void kvm_mmu_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp)
 	}
 	kvm_mmu_page_unlink_children(kvm, sp);
 	if (!sp->root_count) {
-		if (!sp->role.metaphysical)
+		if (!sp->role.metaphysical && !sp->role.invalid)
 			unaccount_shadowed(kvm, sp->gfn);
 		hlist_del(&sp->hash_link);
 		kvm_mmu_free_page(kvm, sp);
 	} else {
+		int invalid = sp->role.invalid;
 		list_move(&sp->link, &kvm->arch.active_mmu_pages);
 		sp->role.invalid = 1;
 		kvm_reload_remote_mmus(kvm);
+		if (!sp->role.metaphysical && !invalid)
+			unaccount_shadowed(kvm, sp->gfn);
 	}
 	kvm_mmu_reset_last_pte_updated(kvm);
 }
@@ -1103,7 +1119,7 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 		mark_page_dirty(vcpu->kvm, gfn);
 
 	pgprintk("%s: setting spte %llx\n", __func__, spte);
-	pgprintk("instantiating %s PTE (%s) at %d (%llx) addr %llx\n",
+	pgprintk("instantiating %s PTE (%s) at %ld (%llx) addr %p\n",
 		 (spte&PT_PAGE_SIZE_MASK)? "2MB" : "4kB",
 		 (spte&PT_WRITABLE_MASK)?"RW":"R", gfn, spte, shadow_pte);
 	set_shadow_pte(shadow_pte, spte);
@@ -1122,8 +1138,10 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 		else
 			kvm_release_pfn_clean(pfn);
 	}
-	if (!ptwrite || !*ptwrite)
+	if (speculative) {
 		vcpu->arch.last_pte_updated = shadow_pte;
+		vcpu->arch.last_pte_gfn = gfn;
+	}
 }
 
 static void nonpaging_new_cr3(struct kvm_vcpu *vcpu)
@@ -1171,9 +1189,10 @@ static int __direct_map(struct kvm_vcpu *vcpu, gpa_t v, int write,
 				return -ENOMEM;
 			}
 
-			table[index] = __pa(new_table->spt)
-				| PT_PRESENT_MASK | PT_WRITABLE_MASK
-				| shadow_user_mask | shadow_x_mask;
+			set_shadow_pte(&table[index],
+				       __pa(new_table->spt)
+				       | PT_PRESENT_MASK | PT_WRITABLE_MASK
+				       | shadow_user_mask | shadow_x_mask);
 		}
 		table_addr = table[index] & PT64_BASE_ADDR_MASK;
 	}
@@ -1211,15 +1230,6 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn)
 }
 
 
-static void nonpaging_prefetch_page(struct kvm_vcpu *vcpu,
-				    struct kvm_mmu_page *sp)
-{
-	int i;
-
-	for (i = 0; i < PT64_ENT_PER_PAGE; ++i)
-		sp->spt[i] = shadow_trap_nonpresent_pte;
-}
-
 static void mmu_free_roots(struct kvm_vcpu *vcpu)
 {
 	int i;
@@ -1671,6 +1681,18 @@ static void mmu_guess_page_from_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	vcpu->arch.update_pte.pfn = pfn;
 }
 
+static void kvm_mmu_access_page(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+	u64 *spte = vcpu->arch.last_pte_updated;
+
+	if (spte
+	    && vcpu->arch.last_pte_gfn == gfn
+	    && shadow_accessed_mask
+	    && !(*spte & shadow_accessed_mask)
+	    && is_shadow_present_pte(*spte))
+		set_bit(PT_ACCESSED_SHIFT, (unsigned long *)spte);
+}
+
 void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 		       const u8 *new, int bytes)
 {
@@ -1694,6 +1716,7 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	pgprintk("%s: gpa %llx bytes %d\n", __func__, gpa, bytes);
 	mmu_guess_page_from_pte_write(vcpu, gpa, new, bytes);
 	spin_lock(&vcpu->kvm->mmu_lock);
+	kvm_mmu_access_page(vcpu, gfn);
 	kvm_mmu_free_some_pages(vcpu);
 	++vcpu->kvm->stat.mmu_pte_write;
 	kvm_mmu_audit(vcpu, "pre pte write");
@@ -1948,7 +1971,7 @@ void kvm_mmu_zap_all(struct kvm *kvm)
 	kvm_flush_remote_tlbs(kvm);
 }
 
-void kvm_mmu_remove_one_alloc_mmu_page(struct kvm *kvm)
+static void kvm_mmu_remove_one_alloc_mmu_page(struct kvm *kvm)
 {
 	struct kvm_mmu_page *page;
 
@@ -1968,6 +1991,8 @@ static int mmu_shrink(int nr_to_scan, gfp_t gfp_mask)
 	list_for_each_entry(kvm, &vm_list, vm_list) {
 		int npages;
 
+		if (!down_read_trylock(&kvm->slots_lock))
+			continue;
 		spin_lock(&kvm->mmu_lock);
 		npages = kvm->arch.n_alloc_mmu_pages -
 			 kvm->arch.n_free_mmu_pages;
@@ -1980,6 +2005,7 @@ static int mmu_shrink(int nr_to_scan, gfp_t gfp_mask)
 		nr_to_scan--;
 
 		spin_unlock(&kvm->mmu_lock);
+		up_read(&kvm->slots_lock);
 	}
 	if (kvm_freed)
 		list_move_tail(&kvm_freed->vm_list, &vm_list);
diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
index 1730757bbc7..258e5d56298 100644
--- a/arch/x86/kvm/mmu.h
+++ b/arch/x86/kvm/mmu.h
@@ -15,7 +15,8 @@
 #define PT_USER_MASK (1ULL << 2)
 #define PT_PWT_MASK (1ULL << 3)
 #define PT_PCD_MASK (1ULL << 4)
-#define PT_ACCESSED_MASK (1ULL << 5)
+#define PT_ACCESSED_SHIFT 5
+#define PT_ACCESSED_MASK (1ULL << PT_ACCESSED_SHIFT)
 #define PT_DIRTY_MASK (1ULL << 6)
 #define PT_PAGE_SIZE_MASK (1ULL << 7)
 #define PT_PAT_MASK (1ULL << 7)
diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index 934c7b61939..4d918220bae 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -460,8 +460,9 @@ static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr)
 static void FNAME(prefetch_page)(struct kvm_vcpu *vcpu,
 				 struct kvm_mmu_page *sp)
 {
-	int i, offset = 0, r = 0;
-	pt_element_t pt;
+	int i, j, offset, r;
+	pt_element_t pt[256 / sizeof(pt_element_t)];
+	gpa_t pte_gpa;
 
 	if (sp->role.metaphysical
 	    || (PTTYPE == 32 && sp->role.level > PT_PAGE_TABLE_LEVEL)) {
@@ -469,19 +470,20 @@ static void FNAME(prefetch_page)(struct kvm_vcpu *vcpu,
 		return;
 	}
 
-	if (PTTYPE == 32)
+	pte_gpa = gfn_to_gpa(sp->gfn);
+	if (PTTYPE == 32) {
 		offset = sp->role.quadrant << PT64_LEVEL_BITS;
+		pte_gpa += offset * sizeof(pt_element_t);
+	}
 
-	for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
-		gpa_t pte_gpa = gfn_to_gpa(sp->gfn);
-		pte_gpa += (i+offset) * sizeof(pt_element_t);
-
-		r = kvm_read_guest_atomic(vcpu->kvm, pte_gpa, &pt,
-					  sizeof(pt_element_t));
-		if (r || is_present_pte(pt))
-			sp->spt[i] = shadow_trap_nonpresent_pte;
-		else
-			sp->spt[i] = shadow_notrap_nonpresent_pte;
+	for (i = 0; i < PT64_ENT_PER_PAGE; i += ARRAY_SIZE(pt)) {
+		r = kvm_read_guest_atomic(vcpu->kvm, pte_gpa, pt, sizeof pt);
+		pte_gpa += ARRAY_SIZE(pt) * sizeof(pt_element_t);
+		for (j = 0; j < ARRAY_SIZE(pt); ++j)
+			if (r || is_present_pte(pt[j]))
+				sp->spt[i+j] = shadow_trap_nonpresent_pte;
+			else
+				sp->spt[i+j] = shadow_notrap_nonpresent_pte;
 	}
 }
 
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 6b0d5fa5bab..b756e876dce 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -27,6 +27,8 @@
 
 #include <asm/desc.h>
 
+#define __ex(x) __kvm_handle_fault_on_reboot(x)
+
 MODULE_AUTHOR("Qumranet");
 MODULE_LICENSE("GPL");
 
@@ -129,17 +131,17 @@ static inline void push_irq(struct kvm_vcpu *vcpu, u8 irq)
 
 static inline void clgi(void)
 {
-	asm volatile (SVM_CLGI);
+	asm volatile (__ex(SVM_CLGI));
 }
 
 static inline void stgi(void)
 {
-	asm volatile (SVM_STGI);
+	asm volatile (__ex(SVM_STGI));
 }
 
 static inline void invlpga(unsigned long addr, u32 asid)
 {
-	asm volatile (SVM_INVLPGA :: "a"(addr), "c"(asid));
+	asm volatile (__ex(SVM_INVLPGA) :: "a"(addr), "c"(asid));
 }
 
 static inline unsigned long kvm_read_cr2(void)
@@ -270,19 +272,11 @@ static int has_svm(void)
 
 static void svm_hardware_disable(void *garbage)
 {
-	struct svm_cpu_data *svm_data
-		= per_cpu(svm_data, raw_smp_processor_id());
-
-	if (svm_data) {
-		uint64_t efer;
+	uint64_t efer;
 
-		wrmsrl(MSR_VM_HSAVE_PA, 0);
-		rdmsrl(MSR_EFER, efer);
-		wrmsrl(MSR_EFER, efer & ~MSR_EFER_SVME_MASK);
-		per_cpu(svm_data, raw_smp_processor_id()) = NULL;
-		__free_page(svm_data->save_area);
-		kfree(svm_data);
-	}
+	wrmsrl(MSR_VM_HSAVE_PA, 0);
+	rdmsrl(MSR_EFER, efer);
+	wrmsrl(MSR_EFER, efer & ~MSR_EFER_SVME_MASK);
 }
 
 static void svm_hardware_enable(void *garbage)
@@ -321,6 +315,19 @@ static void svm_hardware_enable(void *garbage)
 	       page_to_pfn(svm_data->save_area) << PAGE_SHIFT);
 }
 
+static void svm_cpu_uninit(int cpu)
+{
+	struct svm_cpu_data *svm_data
+		= per_cpu(svm_data, raw_smp_processor_id());
+
+	if (!svm_data)
+		return;
+
+	per_cpu(svm_data, raw_smp_processor_id()) = NULL;
+	__free_page(svm_data->save_area);
+	kfree(svm_data);
+}
+
 static int svm_cpu_init(int cpu)
 {
 	struct svm_cpu_data *svm_data;
@@ -458,6 +465,11 @@ err:
 
 static __exit void svm_hardware_unsetup(void)
 {
+	int cpu;
+
+	for_each_online_cpu(cpu)
+		svm_cpu_uninit(cpu);
+
 	__free_pages(pfn_to_page(iopm_base >> PAGE_SHIFT), IOPM_ALLOC_ORDER);
 	iopm_base = 0;
 }
@@ -707,10 +719,6 @@ static void svm_vcpu_put(struct kvm_vcpu *vcpu)
 	rdtscll(vcpu->arch.host_tsc);
 }
 
-static void svm_vcpu_decache(struct kvm_vcpu *vcpu)
-{
-}
-
 static void svm_cache_regs(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
@@ -949,7 +957,9 @@ static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *svm_data)
 
 static unsigned long svm_get_dr(struct kvm_vcpu *vcpu, int dr)
 {
-	return to_svm(vcpu)->db_regs[dr];
+	unsigned long val = to_svm(vcpu)->db_regs[dr];
+	KVMTRACE_2D(DR_READ, vcpu, (u32)dr, (u32)val, handler);
+	return val;
 }
 
 static void svm_set_dr(struct kvm_vcpu *vcpu, int dr, unsigned long value,
@@ -1004,6 +1014,16 @@ static int pf_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 
 	fault_address  = svm->vmcb->control.exit_info_2;
 	error_code = svm->vmcb->control.exit_info_1;
+
+	if (!npt_enabled)
+		KVMTRACE_3D(PAGE_FAULT, &svm->vcpu, error_code,
+			    (u32)fault_address, (u32)(fault_address >> 32),
+			    handler);
+	else
+		KVMTRACE_3D(TDP_FAULT, &svm->vcpu, error_code,
+			    (u32)fault_address, (u32)(fault_address >> 32),
+			    handler);
+
 	return kvm_mmu_page_fault(&svm->vcpu, fault_address, error_code);
 }
 
@@ -1081,6 +1101,19 @@ static int io_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 	return kvm_emulate_pio(&svm->vcpu, kvm_run, in, size, port);
 }
 
+static int nmi_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
+{
+	KVMTRACE_0D(NMI, &svm->vcpu, handler);
+	return 1;
+}
+
+static int intr_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
+{
+	++svm->vcpu.stat.irq_exits;
+	KVMTRACE_0D(INTR, &svm->vcpu, handler);
+	return 1;
+}
+
 static int nop_on_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 {
 	return 1;
@@ -1219,6 +1252,9 @@ static int rdmsr_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 	if (svm_get_msr(&svm->vcpu, ecx, &data))
 		kvm_inject_gp(&svm->vcpu, 0);
 	else {
+		KVMTRACE_3D(MSR_READ, &svm->vcpu, ecx, (u32)data,
+			    (u32)(data >> 32), handler);
+
 		svm->vmcb->save.rax = data & 0xffffffff;
 		svm->vcpu.arch.regs[VCPU_REGS_RDX] = data >> 32;
 		svm->next_rip = svm->vmcb->save.rip + 2;
@@ -1284,16 +1320,19 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, unsigned ecx, u64 data)
 	case MSR_K7_EVNTSEL1:
 	case MSR_K7_EVNTSEL2:
 	case MSR_K7_EVNTSEL3:
+	case MSR_K7_PERFCTR0:
+	case MSR_K7_PERFCTR1:
+	case MSR_K7_PERFCTR2:
+	case MSR_K7_PERFCTR3:
 		/*
-		 * only support writing 0 to the performance counters for now
-		 * to make Windows happy. Should be replaced by a real
-		 * performance counter emulation later.
+		 * Just discard all writes to the performance counters; this
+		 * should keep both older linux and windows 64-bit guests
+		 * happy
 		 */
-		if (data != 0)
-			goto unhandled;
+		pr_unimpl(vcpu, "unimplemented perfctr wrmsr: 0x%x data 0x%llx\n", ecx, data);
+
 		break;
 	default:
-	unhandled:
 		return kvm_set_msr_common(vcpu, ecx, data);
 	}
 	return 0;
@@ -1304,6 +1343,10 @@ static int wrmsr_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 	u32 ecx = svm->vcpu.arch.regs[VCPU_REGS_RCX];
 	u64 data = (svm->vmcb->save.rax & -1u)
 		| ((u64)(svm->vcpu.arch.regs[VCPU_REGS_RDX] & -1u) << 32);
+
+	KVMTRACE_3D(MSR_WRITE, &svm->vcpu, ecx, (u32)data, (u32)(data >> 32),
+		    handler);
+
 	svm->next_rip = svm->vmcb->save.rip + 2;
 	if (svm_set_msr(&svm->vcpu, ecx, data))
 		kvm_inject_gp(&svm->vcpu, 0);
@@ -1323,6 +1366,8 @@ static int msr_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 static int interrupt_window_interception(struct vcpu_svm *svm,
 				   struct kvm_run *kvm_run)
 {
+	KVMTRACE_0D(PEND_INTR, &svm->vcpu, handler);
+
 	svm->vmcb->control.intercept &= ~(1ULL << INTERCEPT_VINTR);
 	svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
 	/*
@@ -1364,8 +1409,8 @@ static int (*svm_exit_handlers[])(struct vcpu_svm *svm,
 	[SVM_EXIT_EXCP_BASE + PF_VECTOR] 	= pf_interception,
 	[SVM_EXIT_EXCP_BASE + NM_VECTOR] 	= nm_interception,
 	[SVM_EXIT_EXCP_BASE + MC_VECTOR] 	= mc_interception,
-	[SVM_EXIT_INTR] 			= nop_on_interception,
-	[SVM_EXIT_NMI]				= nop_on_interception,
+	[SVM_EXIT_INTR] 			= intr_interception,
+	[SVM_EXIT_NMI]				= nmi_interception,
 	[SVM_EXIT_SMI]				= nop_on_interception,
 	[SVM_EXIT_INIT]				= nop_on_interception,
 	[SVM_EXIT_VINTR]			= interrupt_window_interception,
@@ -1397,6 +1442,9 @@ static int handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
 	struct vcpu_svm *svm = to_svm(vcpu);
 	u32 exit_code = svm->vmcb->control.exit_code;
 
+	KVMTRACE_3D(VMEXIT, vcpu, exit_code, (u32)svm->vmcb->save.rip,
+		    (u32)((u64)svm->vmcb->save.rip >> 32), entryexit);
+
 	if (npt_enabled) {
 		int mmu_reload = 0;
 		if ((vcpu->arch.cr0 ^ svm->vmcb->save.cr0) & X86_CR0_PG) {
@@ -1470,6 +1518,8 @@ static inline void svm_inject_irq(struct vcpu_svm *svm, int irq)
 {
 	struct vmcb_control_area *control;
 
+	KVMTRACE_1D(INJ_VIRQ, &svm->vcpu, (u32)irq, handler);
+
 	control = &svm->vmcb->control;
 	control->int_vector = irq;
 	control->int_ctl &= ~V_INTR_PRIO_MASK;
@@ -1660,9 +1710,9 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	sync_lapic_to_cr8(vcpu);
 
 	save_host_msrs(vcpu);
-	fs_selector = read_fs();
-	gs_selector = read_gs();
-	ldt_selector = read_ldt();
+	fs_selector = kvm_read_fs();
+	gs_selector = kvm_read_gs();
+	ldt_selector = kvm_read_ldt();
 	svm->host_cr2 = kvm_read_cr2();
 	svm->host_dr6 = read_dr6();
 	svm->host_dr7 = read_dr7();
@@ -1716,17 +1766,17 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		/* Enter guest mode */
 		"push %%rax \n\t"
 		"mov %c[vmcb](%[svm]), %%rax \n\t"
-		SVM_VMLOAD "\n\t"
-		SVM_VMRUN "\n\t"
-		SVM_VMSAVE "\n\t"
+		__ex(SVM_VMLOAD) "\n\t"
+		__ex(SVM_VMRUN) "\n\t"
+		__ex(SVM_VMSAVE) "\n\t"
 		"pop %%rax \n\t"
 #else
 		/* Enter guest mode */
 		"push %%eax \n\t"
 		"mov %c[vmcb](%[svm]), %%eax \n\t"
-		SVM_VMLOAD "\n\t"
-		SVM_VMRUN "\n\t"
-		SVM_VMSAVE "\n\t"
+		__ex(SVM_VMLOAD) "\n\t"
+		__ex(SVM_VMRUN) "\n\t"
+		__ex(SVM_VMSAVE) "\n\t"
 		"pop %%eax \n\t"
 #endif
 
@@ -1795,9 +1845,9 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	write_dr7(svm->host_dr7);
 	kvm_write_cr2(svm->host_cr2);
 
-	load_fs(fs_selector);
-	load_gs(gs_selector);
-	load_ldt(ldt_selector);
+	kvm_load_fs(fs_selector);
+	kvm_load_gs(gs_selector);
+	kvm_load_ldt(ldt_selector);
 	load_host_msrs(vcpu);
 
 	reload_tss(vcpu);
@@ -1889,7 +1939,6 @@ static struct kvm_x86_ops svm_x86_ops = {
 	.prepare_guest_switch = svm_prepare_guest_switch,
 	.vcpu_load = svm_vcpu_load,
 	.vcpu_put = svm_vcpu_put,
-	.vcpu_decache = svm_vcpu_decache,
 
 	.set_guest_debug = svm_guest_debug,
 	.get_msr = svm_get_msr,
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 10ce6ee4c49..0cac6370171 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -30,6 +30,8 @@
 #include <asm/io.h>
 #include <asm/desc.h>
 
+#define __ex(x) __kvm_handle_fault_on_reboot(x)
+
 MODULE_AUTHOR("Qumranet");
 MODULE_LICENSE("GPL");
 
@@ -53,6 +55,7 @@ struct vmcs {
 
 struct vcpu_vmx {
 	struct kvm_vcpu       vcpu;
+	struct list_head      local_vcpus_link;
 	int                   launched;
 	u8                    fail;
 	u32                   idt_vectoring_info;
@@ -88,9 +91,11 @@ static inline struct vcpu_vmx *to_vmx(struct kvm_vcpu *vcpu)
 }
 
 static int init_rmode(struct kvm *kvm);
+static u64 construct_eptp(unsigned long root_hpa);
 
 static DEFINE_PER_CPU(struct vmcs *, vmxarea);
 static DEFINE_PER_CPU(struct vmcs *, current_vmcs);
+static DEFINE_PER_CPU(struct list_head, vcpus_on_cpu);
 
 static struct page *vmx_io_bitmap_a;
 static struct page *vmx_io_bitmap_b;
@@ -260,6 +265,11 @@ static inline int cpu_has_vmx_vpid(void)
 		SECONDARY_EXEC_ENABLE_VPID);
 }
 
+static inline int cpu_has_virtual_nmis(void)
+{
+	return vmcs_config.pin_based_exec_ctrl & PIN_BASED_VIRTUAL_NMIS;
+}
+
 static int __find_msr_index(struct vcpu_vmx *vmx, u32 msr)
 {
 	int i;
@@ -278,7 +288,7 @@ static inline void __invvpid(int ext, u16 vpid, gva_t gva)
 	u64 gva;
     } operand = { vpid, 0, gva };
 
-    asm volatile (ASM_VMX_INVVPID
+    asm volatile (__ex(ASM_VMX_INVVPID)
 		  /* CF==1 or ZF==1 --> rc = -1 */
 		  "; ja 1f ; ud2 ; 1:"
 		  : : "a"(&operand), "c"(ext) : "cc", "memory");
@@ -290,7 +300,7 @@ static inline void __invept(int ext, u64 eptp, gpa_t gpa)
 		u64 eptp, gpa;
 	} operand = {eptp, gpa};
 
-	asm volatile (ASM_VMX_INVEPT
+	asm volatile (__ex(ASM_VMX_INVEPT)
 			/* CF==1 or ZF==1 --> rc = -1 */
 			"; ja 1f ; ud2 ; 1:\n"
 			: : "a" (&operand), "c" (ext) : "cc", "memory");
@@ -311,7 +321,7 @@ static void vmcs_clear(struct vmcs *vmcs)
 	u64 phys_addr = __pa(vmcs);
 	u8 error;
 
-	asm volatile (ASM_VMX_VMCLEAR_RAX "; setna %0"
+	asm volatile (__ex(ASM_VMX_VMCLEAR_RAX) "; setna %0"
 		      : "=g"(error) : "a"(&phys_addr), "m"(phys_addr)
 		      : "cc", "memory");
 	if (error)
@@ -329,6 +339,9 @@ static void __vcpu_clear(void *arg)
 	if (per_cpu(current_vmcs, cpu) == vmx->vmcs)
 		per_cpu(current_vmcs, cpu) = NULL;
 	rdtscll(vmx->vcpu.arch.host_tsc);
+	list_del(&vmx->local_vcpus_link);
+	vmx->vcpu.cpu = -1;
+	vmx->launched = 0;
 }
 
 static void vcpu_clear(struct vcpu_vmx *vmx)
@@ -336,7 +349,6 @@ static void vcpu_clear(struct vcpu_vmx *vmx)
 	if (vmx->vcpu.cpu == -1)
 		return;
 	smp_call_function_single(vmx->vcpu.cpu, __vcpu_clear, vmx, 1);
-	vmx->launched = 0;
 }
 
 static inline void vpid_sync_vcpu_all(struct vcpu_vmx *vmx)
@@ -378,7 +390,7 @@ static unsigned long vmcs_readl(unsigned long field)
 {
 	unsigned long value;
 
-	asm volatile (ASM_VMX_VMREAD_RDX_RAX
+	asm volatile (__ex(ASM_VMX_VMREAD_RDX_RAX)
 		      : "=a"(value) : "d"(field) : "cc");
 	return value;
 }
@@ -413,7 +425,7 @@ static void vmcs_writel(unsigned long field, unsigned long value)
 {
 	u8 error;
 
-	asm volatile (ASM_VMX_VMWRITE_RAX_RDX "; setna %0"
+	asm volatile (__ex(ASM_VMX_VMWRITE_RAX_RDX) "; setna %0"
 		       : "=q"(error) : "a"(value), "d"(field) : "cc");
 	if (unlikely(error))
 		vmwrite_error(field, value);
@@ -431,10 +443,8 @@ static void vmcs_write32(unsigned long field, u32 value)
 
 static void vmcs_write64(unsigned long field, u64 value)
 {
-#ifdef CONFIG_X86_64
-	vmcs_writel(field, value);
-#else
 	vmcs_writel(field, value);
+#ifndef CONFIG_X86_64
 	asm volatile ("");
 	vmcs_writel(field+1, value >> 32);
 #endif
@@ -474,7 +484,7 @@ static void reload_tss(void)
 	struct descriptor_table gdt;
 	struct desc_struct *descs;
 
-	get_gdt(&gdt);
+	kvm_get_gdt(&gdt);
 	descs = (void *)gdt.base;
 	descs[GDT_ENTRY_TSS].type = 9; /* available TSS */
 	load_TR_desc();
@@ -530,9 +540,9 @@ static void vmx_save_host_state(struct kvm_vcpu *vcpu)
 	 * Set host fs and gs selectors.  Unfortunately, 22.2.3 does not
 	 * allow segment selectors with cpl > 0 or ti == 1.
 	 */
-	vmx->host_state.ldt_sel = read_ldt();
+	vmx->host_state.ldt_sel = kvm_read_ldt();
 	vmx->host_state.gs_ldt_reload_needed = vmx->host_state.ldt_sel;
-	vmx->host_state.fs_sel = read_fs();
+	vmx->host_state.fs_sel = kvm_read_fs();
 	if (!(vmx->host_state.fs_sel & 7)) {
 		vmcs_write16(HOST_FS_SELECTOR, vmx->host_state.fs_sel);
 		vmx->host_state.fs_reload_needed = 0;
@@ -540,7 +550,7 @@ static void vmx_save_host_state(struct kvm_vcpu *vcpu)
 		vmcs_write16(HOST_FS_SELECTOR, 0);
 		vmx->host_state.fs_reload_needed = 1;
 	}
-	vmx->host_state.gs_sel = read_gs();
+	vmx->host_state.gs_sel = kvm_read_gs();
 	if (!(vmx->host_state.gs_sel & 7))
 		vmcs_write16(HOST_GS_SELECTOR, vmx->host_state.gs_sel);
 	else {
@@ -576,15 +586,15 @@ static void __vmx_load_host_state(struct vcpu_vmx *vmx)
 	++vmx->vcpu.stat.host_state_reload;
 	vmx->host_state.loaded = 0;
 	if (vmx->host_state.fs_reload_needed)
-		load_fs(vmx->host_state.fs_sel);
+		kvm_load_fs(vmx->host_state.fs_sel);
 	if (vmx->host_state.gs_ldt_reload_needed) {
-		load_ldt(vmx->host_state.ldt_sel);
+		kvm_load_ldt(vmx->host_state.ldt_sel);
 		/*
 		 * If we have to reload gs, we must take care to
 		 * preserve our gs base.
 		 */
 		local_irq_save(flags);
-		load_gs(vmx->host_state.gs_sel);
+		kvm_load_gs(vmx->host_state.gs_sel);
 #ifdef CONFIG_X86_64
 		wrmsrl(MSR_GS_BASE, vmcs_readl(HOST_GS_BASE));
 #endif
@@ -617,13 +627,17 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 		vcpu_clear(vmx);
 		kvm_migrate_timers(vcpu);
 		vpid_sync_vcpu_all(vmx);
+		local_irq_disable();
+		list_add(&vmx->local_vcpus_link,
+			 &per_cpu(vcpus_on_cpu, cpu));
+		local_irq_enable();
 	}
 
 	if (per_cpu(current_vmcs, cpu) != vmx->vmcs) {
 		u8 error;
 
 		per_cpu(current_vmcs, cpu) = vmx->vmcs;
-		asm volatile (ASM_VMX_VMPTRLD_RAX "; setna %0"
+		asm volatile (__ex(ASM_VMX_VMPTRLD_RAX) "; setna %0"
 			      : "=g"(error) : "a"(&phys_addr), "m"(phys_addr)
 			      : "cc");
 		if (error)
@@ -640,8 +654,8 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 		 * Linux uses per-cpu TSS and GDT, so set these when switching
 		 * processors.
 		 */
-		vmcs_writel(HOST_TR_BASE, read_tr_base()); /* 22.2.4 */
-		get_gdt(&dt);
+		vmcs_writel(HOST_TR_BASE, kvm_read_tr_base()); /* 22.2.4 */
+		kvm_get_gdt(&dt);
 		vmcs_writel(HOST_GDTR_BASE, dt.base);   /* 22.2.4 */
 
 		rdmsrl(MSR_IA32_SYSENTER_ESP, sysenter_esp);
@@ -684,11 +698,6 @@ static void vmx_fpu_deactivate(struct kvm_vcpu *vcpu)
 	update_exception_bitmap(vcpu);
 }
 
-static void vmx_vcpu_decache(struct kvm_vcpu *vcpu)
-{
-	vcpu_clear(to_vmx(vcpu));
-}
-
 static unsigned long vmx_get_rflags(struct kvm_vcpu *vcp