1 files changed, 527 insertions, 108 deletions

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 1eefebe5d727..c243b81e3c74 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -46,6 +46,8 @@
 #include <linux/uaccess.h>
 #include <linux/hash.h>
 #include <linux/pci.h>
+#include <linux/timekeeper_internal.h>
+#include <linux/pvclock_gtod.h>
 #include <trace/events/kvm.h>
 
 #define CREATE_TRACE_POINTS
@@ -118,7 +120,7 @@ struct kvm_shared_msrs {
 };
 
 static struct kvm_shared_msrs_global __read_mostly shared_msrs_global;
-static DEFINE_PER_CPU(struct kvm_shared_msrs, shared_msrs);
+static struct kvm_shared_msrs __percpu *shared_msrs;
 
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "pf_fixed", VCPU_STAT(pf_fixed) },
@@ -158,7 +160,9 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 
 u64 __read_mostly host_xcr0;
 
-int emulator_fix_hypercall(struct x86_emulate_ctxt *ctxt);
+static int emulator_fix_hypercall(struct x86_emulate_ctxt *ctxt);
+
+static int kvm_vcpu_reset(struct kvm_vcpu *vcpu);
 
 static inline void kvm_async_pf_hash_reset(struct kvm_vcpu *vcpu)
 {
@@ -187,10 +191,10 @@ static void kvm_on_user_return(struct user_return_notifier *urn)
 
 static void shared_msr_update(unsigned slot, u32 msr)
 {
-	struct kvm_shared_msrs *smsr;
 	u64 value;
+	unsigned int cpu = smp_processor_id();
+	struct kvm_shared_msrs *smsr = per_cpu_ptr(shared_msrs, cpu);
 
-	smsr = &__get_cpu_var(shared_msrs);
 	/* only read, and nobody should modify it at this time,
 	 * so don't need lock */
 	if (slot >= shared_msrs_global.nr) {
@@ -222,7 +226,8 @@ static void kvm_shared_msr_cpu_online(void)
 
 void kvm_set_shared_msr(unsigned slot, u64 value, u64 mask)
 {
-	struct kvm_shared_msrs *smsr = &__get_cpu_var(shared_msrs);
+	unsigned int cpu = smp_processor_id();
+	struct kvm_shared_msrs *smsr = per_cpu_ptr(shared_msrs, cpu);
 
 	if (((value ^ smsr->values[slot].curr) & mask) == 0)
 		return;
@@ -238,7 +243,8 @@ EXPORT_SYMBOL_GPL(kvm_set_shared_msr);
 
 static void drop_user_return_notifiers(void *ignore)
 {
-	struct kvm_shared_msrs *smsr = &__get_cpu_var(shared_msrs);
+	unsigned int cpu = smp_processor_id();
+	struct kvm_shared_msrs *smsr = per_cpu_ptr(shared_msrs, cpu);
 
 	if (smsr->registered)
 		kvm_on_user_return(&smsr->urn);
@@ -633,7 +639,7 @@ int kvm_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
 	}
 
 	if (is_long_mode(vcpu)) {
-		if (kvm_read_cr4(vcpu) & X86_CR4_PCIDE) {
+		if (kvm_read_cr4_bits(vcpu, X86_CR4_PCIDE)) {
 			if (cr3 & CR3_PCID_ENABLED_RESERVED_BITS)
 				return 1;
 		} else
@@ -827,6 +833,7 @@ static u32 msrs_to_save[] = {
 static unsigned num_msrs_to_save;
 
 static const u32 emulated_msrs[] = {
+	MSR_IA32_TSC_ADJUST,
 	MSR_IA32_TSCDEADLINE,
 	MSR_IA32_MISC_ENABLE,
 	MSR_IA32_MCG_STATUS,
@@ -886,9 +893,9 @@ EXPORT_SYMBOL_GPL(kvm_enable_efer_bits);
  * Returns 0 on success, non-0 otherwise.
  * Assumes vcpu_load() was already called.
  */
-int kvm_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)
+int kvm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
 {
-	return kvm_x86_ops->set_msr(vcpu, msr_index, data);
+	return kvm_x86_ops->set_msr(vcpu, msr);
 }
 
 /*
@@ -896,8 +903,62 @@ int kvm_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)
  */
 static int do_set_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
 {
-	return kvm_set_msr(vcpu, index, *data);
+	struct msr_data msr;
+
+	msr.data = *data;
+	msr.index = index;
+	msr.host_initiated = true;
+	return kvm_set_msr(vcpu, &msr);
+}
+
+#ifdef CONFIG_X86_64
+struct pvclock_gtod_data {
+	seqcount_t	seq;
+
+	struct { /* extract of a clocksource struct */
+		int vclock_mode;
+		cycle_t	cycle_last;
+		cycle_t	mask;
+		u32	mult;
+		u32	shift;
+	} clock;
+
+	/* open coded 'struct timespec' */
+	u64		monotonic_time_snsec;
+	time_t		monotonic_time_sec;
+};
+
+static struct pvclock_gtod_data pvclock_gtod_data;
+
+static void update_pvclock_gtod(struct timekeeper *tk)
+{
+	struct pvclock_gtod_data *vdata = &pvclock_gtod_data;
+
+	write_seqcount_begin(&vdata->seq);
+
+	/* copy pvclock gtod data */
+	vdata->clock.vclock_mode	= tk->clock->archdata.vclock_mode;
+	vdata->clock.cycle_last		= tk->clock->cycle_last;
+	vdata->clock.mask		= tk->clock->mask;
+	vdata->clock.mult		= tk->mult;
+	vdata->clock.shift		= tk->shift;
+
+	vdata->monotonic_time_sec	= tk->xtime_sec
+					+ tk->wall_to_monotonic.tv_sec;
+	vdata->monotonic_time_snsec	= tk->xtime_nsec
+					+ (tk->wall_to_monotonic.tv_nsec
+						<< tk->shift);
+	while (vdata->monotonic_time_snsec >=
+					(((u64)NSEC_PER_SEC) << tk->shift)) {
+		vdata->monotonic_time_snsec -=
+					((u64)NSEC_PER_SEC) << tk->shift;
+		vdata->monotonic_time_sec++;
+	}
+
+	write_seqcount_end(&vdata->seq);
 }
+#endif
+
 
 static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock)
 {
@@ -995,6 +1056,10 @@ static inline u64 get_kernel_ns(void)
 	return timespec_to_ns(&ts);
 }
 
+#ifdef CONFIG_X86_64
+static atomic_t kvm_guest_has_master_clock = ATOMIC_INIT(0);
+#endif
+
 static DEFINE_PER_CPU(unsigned long, cpu_tsc_khz);
 unsigned long max_tsc_khz;
 
@@ -1046,12 +1111,47 @@ static u64 compute_guest_tsc(struct kvm_vcpu *vcpu, s64 kernel_ns)
 	return tsc;
 }
 
-void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
+void kvm_track_tsc_matching(struct kvm_vcpu *vcpu)
+{
+#ifdef CONFIG_X86_64
+	bool vcpus_matched;
+	bool do_request = false;
+	struct kvm_arch *ka = &vcpu->kvm->arch;
+	struct pvclock_gtod_data *gtod = &pvclock_gtod_data;
+
+	vcpus_matched = (ka->nr_vcpus_matched_tsc + 1 ==
+			 atomic_read(&vcpu->kvm->online_vcpus));
+
+	if (vcpus_matched && gtod->clock.vclock_mode == VCLOCK_TSC)
+		if (!ka->use_master_clock)
+			do_request = 1;
+
+	if (!vcpus_matched && ka->use_master_clock)
+			do_request = 1;
+
+	if (do_request)
+		kvm_make_request(KVM_REQ_MASTERCLOCK_UPDATE, vcpu);
+
+	trace_kvm_track_tsc(vcpu->vcpu_id, ka->nr_vcpus_matched_tsc,
+			    atomic_read(&vcpu->kvm->online_vcpus),
+		            ka->use_master_clock, gtod->clock.vclock_mode);
+#endif
+}
+
+static void update_ia32_tsc_adjust_msr(struct kvm_vcpu *vcpu, s64 offset)
+{
+	u64 curr_offset = kvm_x86_ops->read_tsc_offset(vcpu);
+	vcpu->arch.ia32_tsc_adjust_msr += offset - curr_offset;
+}
+
+void kvm_write_tsc(struct kvm_vcpu *vcpu, struct msr_data *msr)
 {
 	struct kvm *kvm = vcpu->kvm;
 	u64 offset, ns, elapsed;
 	unsigned long flags;
 	s64 usdiff;
+	bool matched;
+	u64 data = msr->data;
 
 	raw_spin_lock_irqsave(&kvm->arch.tsc_write_lock, flags);
 	offset = kvm_x86_ops->compute_tsc_offset(vcpu, data);
@@ -1094,6 +1194,7 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 			offset = kvm_x86_ops->compute_tsc_offset(vcpu, data);
 			pr_debug("kvm: adjusted tsc offset by %llu\n", delta);
 		}
+		matched = true;
 	} else {
 		/*
 		 * We split periods of matched TSC writes into generations.
@@ -1108,6 +1209,7 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 		kvm->arch.cur_tsc_nsec = ns;
 		kvm->arch.cur_tsc_write = data;
 		kvm->arch.cur_tsc_offset = offset;
+		matched = false;
 		pr_debug("kvm: new tsc generation %u, clock %llu\n",
 			 kvm->arch.cur_tsc_generation, data);
 	}
@@ -1129,26 +1231,195 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 	vcpu->arch.this_tsc_nsec = kvm->arch.cur_tsc_nsec;
 	vcpu->arch.this_tsc_write = kvm->arch.cur_tsc_write;
 
+	if (guest_cpuid_has_tsc_adjust(vcpu) && !msr->host_initiated)
+		update_ia32_tsc_adjust_msr(vcpu, offset);
 	kvm_x86_ops->write_tsc_offset(vcpu, offset);
 	raw_spin_unlock_irqrestore(&kvm->arch.tsc_write_lock, flags);
+
+	spin_lock(&kvm->arch.pvclock_gtod_sync_lock);
+	if (matched)
+		kvm->arch.nr_vcpus_matched_tsc++;
+	else
+		kvm->arch.nr_vcpus_matched_tsc = 0;
+
+	kvm_track_tsc_matching(vcpu);
+	spin_unlock(&kvm->arch.pvclock_gtod_sync_lock);
 }
 
 EXPORT_SYMBOL_GPL(kvm_write_tsc);
 
+#ifdef CONFIG_X86_64
+
+static cycle_t read_tsc(void)
+{
+	cycle_t ret;
+	u64 last;
+
+	/*
+	 * Empirically, a fence (of type that depends on the CPU)
+	 * before rdtsc is enough to ensure that rdtsc is ordered
+	 * with respect to loads.  The various CPU manuals are unclear
+	 * as to whether rdtsc can be reordered with later loads,
+	 * but no one has ever seen it happen.
+	 */
+	rdtsc_barrier();
+	ret = (cycle_t)vget_cycles();
+
+	last = pvclock_gtod_data.clock.cycle_last;
+
+	if (likely(ret >= last))
+		return ret;
+
+	/*
+	 * GCC likes to generate cmov here, but this branch is extremely
+	 * predictable (it's just a funciton of time and the likely is
+	 * very likely) and there's a data dependence, so force GCC
+	 * to generate a branch instead.  I don't barrier() because
+	 * we don't actually need a barrier, and if this function
+	 * ever gets inlined it will generate worse code.
+	 */
+	asm volatile ("");
+	return last;
+}
+
+static inline u64 vgettsc(cycle_t *cycle_now)
+{
+	long v;
+	struct pvclock_gtod_data *gtod = &pvclock_gtod_data;
+
+	*cycle_now = read_tsc();
+
+	v = (*cycle_now - gtod->clock.cycle_last) & gtod->clock.mask;
+	return v * gtod->clock.mult;
+}
+
+static int do_monotonic(struct timespec *ts, cycle_t *cycle_now)
+{
+	unsigned long seq;
+	u64 ns;
+	int mode;
+	struct pvclock_gtod_data *gtod = &pvclock_gtod_data;
+
+	ts->tv_nsec = 0;
+	do {
+		seq = read_seqcount_begin(&gtod->seq);
+		mode = gtod->clock.vclock_mode;
+		ts->tv_sec = gtod->monotonic_time_sec;
+		ns = gtod->monotonic_time_snsec;
+		ns += vgettsc(cycle_now);
+		ns >>= gtod->clock.shift;
+	} while (unlikely(read_seqcount_retry(&gtod->seq, seq)));
+	timespec_add_ns(ts, ns);
+
+	return mode;
+}
+
+/* returns true if host is using tsc clocksource */
+static bool kvm_get_time_and_clockread(s64 *kernel_ns, cycle_t *cycle_now)
+{
+	struct timespec ts;
+
+	/* checked again under seqlock below */
+	if (pvclock_gtod_data.clock.vclock_mode != VCLOCK_TSC)
+		return false;
+
+	if (do_monotonic(&ts, cycle_now) != VCLOCK_TSC)
+		return false;
+
+	monotonic_to_bootbased(&ts);
+	*kernel_ns = timespec_to_ns(&ts);
+
+	return true;
+}
+#endif
+
+/*
+ *
+ * Assuming a stable TSC across physical CPUS, and a stable TSC
+ * across virtual CPUs, the following condition is possible.
+ * Each numbered line represents an event visible to both
+ * CPUs at the next numbered event.
+ *
+ * "timespecX" represents host monotonic time. "tscX" represents
+ * RDTSC value.
+ *
+ * 		VCPU0 on CPU0		|	VCPU1 on CPU1
+ *
+ * 1.  read timespec0,tsc0
+ * 2.					| timespec1 = timespec0 + N
+ * 					| tsc1 = tsc0 + M
+ * 3. transition to guest		| transition to guest
+ * 4. ret0 = timespec0 + (rdtsc - tsc0) |
+ * 5.				        | ret1 = timespec1 + (rdtsc - tsc1)
+ * 				        | ret1 = timespec0 + N + (rdtsc - (tsc0 + M))
+ *
+ * Since ret0 update is visible to VCPU1 at time 5, to obey monotonicity:
+ *
+ * 	- ret0 < ret1
+ *	- timespec0 + (rdtsc - tsc0) < timespec0 + N + (rdtsc - (tsc0 + M))
+ *		...
+ *	- 0 < N - M => M < N
+ *
+ * That is, when timespec0 != timespec1, M < N. Unfortunately that is not
+ * always the case (the difference between two distinct xtime instances
+ * might be smaller then the difference between corresponding TSC reads,
+ * when updating guest vcpus pvclock areas).
+ *
+ * To avoid that problem, do not allow visibility of distinct
+ * system_timestamp/tsc_timestamp values simultaneously: use a master
+ * copy of host monotonic time values. Update that master copy
+ * in lockstep.
+ *
+ * Rely on synchronization of host TSCs and guest TSCs for monotonicity.
+ *
+ */
+
+static void pvclock_update_vm_gtod_copy(struct kvm *kvm)
+{
+#ifdef CONFIG_X86_64
+	struct kvm_arch *ka = &kvm->arch;
+	int vclock_mode;
+	bool host_tsc_clocksource, vcpus_matched;
+
+	vcpus_matched = (ka->nr_vcpus_matched_tsc + 1 ==
+			atomic_read(&kvm->online_vcpus));
+
+	/*
+	 * If the host uses TSC clock, then passthrough TSC as stable
+	 * to the guest.
+	 */
+	host_tsc_clocksource = kvm_get_time_and_clockread(
+					&ka->master_kernel_ns,
+					&ka->master_cycle_now);
+
+	ka->use_master_clock = host_tsc_clocksource & vcpus_matched;
+
+	if (ka->use_master_clock)
+		atomic_set(&kvm_guest_has_master_clock, 1);
+
+	vclock_mode = pvclock_gtod_data.clock.vclock_mode;
+	trace_kvm_update_master_clock(ka->use_master_clock, vclock_mode,
+					vcpus_matched);
+#endif
+}
+
 static int kvm_guest_time_update(struct kvm_vcpu *v)
 {
-	unsigned long flags;
+	unsigned long flags, this_tsc_khz;
 	struct kvm_vcpu_arch *vcpu = &v->arch;
+	struct kvm_arch *ka = &v->kvm->arch;
 	void *shared_kaddr;
-	unsigned long this_tsc_khz;
 	s64 kernel_ns, max_kernel_ns;
-	u64 tsc_timestamp;
+	u64 tsc_timestamp, host_tsc;
+	struct pvclock_vcpu_time_info *guest_hv_clock;
 	u8 pvclock_flags;
+	bool use_master_clock;
+
+	kernel_ns = 0;
+	host_tsc = 0;
 
 	/* Keep irq disabled to prevent changes to the clock */
 	local_irq_save(flags);
-	tsc_timestamp = kvm_x86_ops->read_l1_tsc(v);
-	kernel_ns = get_kernel_ns();
 	this_tsc_khz = __get_cpu_var(cpu_tsc_khz);
 	if (unlikely(this_tsc_khz == 0)) {
 		local_irq_restore(flags);
@@ -1157,6 +1428,24 @@ static int kvm_guest_time_update(struct kvm_vcpu *v)
 	}
 
 	/*
+	 * If the host uses TSC clock, then passthrough TSC as stable
+	 * to the guest.
+	 */
+	spin_lock(&ka->pvclock_gtod_sync_lock);
+	use_master_clock = ka->use_master_clock;
+	if (use_master_clock) {
+		host_tsc = ka->master_cycle_now;
+		kernel_ns = ka->master_kernel_ns;
+	}
+	spin_unlock(&ka->pvclock_gtod_sync_lock);
+	if (!use_master_clock) {
+		host_tsc = native_read_tsc();
+		kernel_ns = get_kernel_ns();
+	}
+
+	tsc_timestamp = kvm_x86_ops->read_l1_tsc(v, host_tsc);
+
+	/*
 	 * We may have to catch up the TSC to match elapsed wall clock
 	 * time for two reasons, even if kvmclock is used.
 	 *   1) CPU could have been running below the maximum TSC rate
@@ -1217,23 +1506,20 @@ static int kvm_guest_time_update(struct kvm_vcpu *v)
 		vcpu->hw_tsc_khz = this_tsc_khz;
 	}
 
-	if (max_kernel_ns > kernel_ns)
-		kernel_ns = max_kernel_ns;
-
+	/* with a master <monotonic time, tsc value> tuple,
+	 * pvclock clock reads always increase at the (scaled) rate
+	 * of guest TSC - no need to deal with sampling errors.
+	 */
+	if (!use_master_clock) {
+		if (max_kernel_ns > kernel_ns)
+			kernel_ns = max_kernel_ns;
+	}
 	/* With all the info we got, fill in the values */
 	vcpu->hv_clock.tsc_timestamp = tsc_timestamp;
 	vcpu->hv_clock.system_time = kernel_ns + v->kvm->arch.kvmclock_offset;
 	vcpu->last_kernel_ns = kernel_ns;
 	vcpu->last_guest_tsc = tsc_timestamp;
 
-	pvclock_flags = 0;
-	if (vcpu->pvclock_set_guest_stopped_request) {
-		pvclock_flags |= PVCLOCK_GUEST_STOPPED;
-		vcpu->pvclock_set_guest_stopped_request = false;
-	}
-
-	vcpu->hv_clock.flags = pvclock_flags;
-
 	/*
 	 * The interface expects us to write an even number signaling that the
 	 * update is finished. Since the guest won't see the intermediate
@@ -1243,6 +1529,22 @@ static int kvm_guest_time_update(struct kvm_vcpu *v)
 
 	shared_kaddr = kmap_atomic(vcpu->time_page);
 
+	guest_hv_clock = shared_kaddr + vcpu->time_offset;
+
+	/* retain PVCLOCK_GUEST_STOPPED if set in guest copy */
+	pvclock_flags = (guest_hv_clock->flags & PVCLOCK_GUEST_STOPPED);
+
+	if (vcpu->pvclock_set_guest_stopped_request) {
+		pvclock_flags |= PVCLOCK_GUEST_STOPPED;
+		vcpu->pvclock_set_guest_stopped_request = false;
+	}
+
+	/* If the host uses TSC clocksource, then it is stable */
+	if (use_master_clock)
+		pvclock_flags |= PVCLOCK_TSC_STABLE_BIT;
+
+	vcpu->hv_clock.flags = pvclock_flags;
+
 	memcpy(shared_kaddr + vcpu->time_offset, &vcpu->hv_clock,
 	       sizeof(vcpu->hv_clock));
 
@@ -1572,9 +1874,11 @@ static void record_steal_time(struct kvm_vcpu *vcpu)
 		&vcpu->arch.st.steal, sizeof(struct kvm_steal_time));
 }
 
-int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
+int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 {
 	bool pr = false;
+	u32 msr = msr_info->index;
+	u64 data = msr_info->data;
 
 	switch (msr) {
 	case MSR_EFER:
@@ -1625,6 +1929,15 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 	case MSR_IA32_TSCDEADLINE:
 		kvm_set_lapic_tscdeadline_msr(vcpu, data);
 		break;
+	case MSR_IA32_TSC_ADJUST:
+		if (guest_cpuid_has_tsc_adjust(vcpu)) {
+			if (!msr_info->host_initiated) {
+				u64 adj = data - vcpu->arch.ia32_tsc_adjust_msr;
+				kvm_x86_ops->adjust_tsc_offset(vcpu, adj, true);
+			}
+			vcpu->arch.ia32_tsc_adjust_msr = data;
+		}
+		break;
 	case MSR_IA32_MISC_ENABLE:
 		vcpu->arch.ia32_misc_enable_msr = data;
 		break;
@@ -1984,6 +2297,9 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 	case MSR_IA32_TSCDEADLINE:
 		data = kvm_get_lapic_tscdeadline_msr(vcpu);
 		break;
+	case MSR_IA32_TSC_ADJUST:
+		data = (u64)vcpu->arch.ia32_tsc_adjust_msr;
+		break;
 	case MSR_IA32_MISC_ENABLE:
 		data = vcpu->arch.ia32_misc_enable_msr;
 		break;
@@ -2342,7 +2658,12 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 			kvm_x86_ops->write_tsc_offset(vcpu, offset);
 			vcpu->arch.tsc_catchup = 1;
 		}
-		kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
+		/*
+		 * On a host with synchronized TSC, there is no need to update
+		 * kvmclock on vcpu->cpu migration
+		 */
+		if (!vcpu->kvm->arch.use_master_clock || vcpu->cpu == -1)
+			kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
 		if (vcpu->cpu != cpu)
 			kvm_migrate_timers(vcpu);
 		vcpu->cpu = cpu;
@@ -2691,15 +3012,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (!vcpu->arch.apic)
 			goto out;
 		u.lapic = memdup_user(argp, sizeof(*u.lapic));
-		if (IS_ERR(u.lapic)) {
-			r = PTR_ERR(u.lapic);
-			goto out;
-		}
+		if (IS_ERR(u.lapic))
+			return PTR_ERR(u.lapic);
 
 		r = kvm_vcpu_ioctl_set_lapic(vcpu, u.lapic);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_INTERRUPT: {
@@ -2709,16 +3025,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (copy_from_user(&irq, argp, sizeof irq))
 			goto out;
 		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_NMI: {
 		r = kvm_vcpu_ioctl_nmi(vcpu);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_SET_CPUID: {
@@ -2729,8 +3039,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
 			goto out;
 		r = kvm_vcpu_ioctl_set_cpuid(vcpu, &cpuid, cpuid_arg->entries);
-		if (r)
-			goto out;
 		break;
 	}
 	case KVM_SET_CPUID2: {
@@ -2742,8 +3050,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 			goto out;
 		r = kvm_vcpu_ioctl_set_cpuid2(vcpu, &cpuid,
 					      cpuid_arg->entries);
-		if (r)
-			goto out;
 		break;
 	}
 	case KVM_GET_CPUID2: {
@@ -2875,10 +3181,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XSAVE: {
 		u.xsave = memdup_user(argp, sizeof(*u.xsave));
-		if (IS_ERR(u.xsave)) {
-			r = PTR_ERR(u.xsave);
-			goto out;
-		}
+		if (IS_ERR(u.xsave))
+			return PTR_ERR(u.xsave);
 
 		r = kvm_vcpu_ioctl_x86_set_xsave(vcpu, u.xsave);
 		break;
@@ -2900,10 +3204,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XCRS: {
 		u.xcrs = memdup_user(argp, sizeof(*u.xcrs));
-		if (IS_ERR(u.xcrs)) {
-			r = PTR_ERR(u.xcrs);
-			goto out;
-		}
+		if (IS_ERR(u.xcrs))
+			return PTR_ERR(u.xcrs);
 
 		r = kvm_vcpu_ioctl_x86_set_xcrs(vcpu, u.xcrs);
 		break;
@@ -2951,7 +3253,7 @@ static int kvm_vm_ioctl_set_tss_addr(struct kvm *kvm, unsigned long addr)
 	int ret;
 
 	if (addr > (unsigned int)(-3 * PAGE_SIZE))
-		return -1;
+		return -EINVAL;
 	ret = kvm_x86_ops->set_tss_addr(kvm, addr);
 	return ret;
 }
@@ -3212,8 +3514,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 	switch (ioctl) {
 	case KVM_SET_TSS_ADDR:
 		r = kvm_vm_ioctl_set_tss_addr(kvm, arg);
-		if (r < 0)
-			goto out;
 		break;
 	case KVM_SET_IDENTITY_MAP_ADDR: {
 		u64 ident_addr;
@@ -3222,14 +3522,10 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		if (copy_from_user(&ident_addr, argp, sizeof ident_addr))
 			goto out;
 		r = kvm_vm_ioctl_set_identity_map_addr(kvm, ident_addr);
-		if (r < 0)
-			goto out;
 		break;
 	}
 	case KVM_SET_NR_MMU_PAGES:
 		r = kvm_vm_ioctl_set_nr_mmu_pages(kvm, arg);
-		if (r)
-			goto out;
 		break;
 	case KVM_GET_NR_MMU_PAGES:
 		r = kvm_vm_ioctl_get_nr_mmu_pages(kvm);
@@ -3320,8 +3616,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		r = 0;
 	get_irqchip_out:
 		kfree(chip);
-		if (r)
-			goto out;
 		break;
 	}
 	case KVM_SET_IRQCHIP: {
@@ -3343,8 +3637,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		r = 0;
 	set_irqchip_out:
 		kfree(chip);
-		if (r)
-			goto out;
 		break;
 	}
 	case KVM_GET_PIT: {
@@ -3371,9 +3663,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		if (!kvm->arch.vpit)
 			goto out;
 		r = kvm_vm_ioctl_set_pit(kvm, &u.ps);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_GET_PIT2: {
@@ -3397,9 +3686,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		if (!kvm->arch.vpit)
 			goto out;
 		r = kvm_vm_ioctl_set_pit2(kvm, &u.ps2);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_REINJECT_CONTROL: {
@@ -3408,9 +3694,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		if (copy_from_user(&control, argp, sizeof(control)))
 			goto out;
 		r = kvm_vm_ioctl_reinject(kvm, &control);
-		if (r)
-			goto out;
-		r = 0;
 		break;
 	}
 	case KVM_XEN_HVM_CONFIG: {
@@ -3779,7 +4062,7 @@ static int write_exit_mmio(struct kvm_vcpu *vcpu, gpa_t gpa,
 {
 	struct kvm_mmio_fragment *frag = &vcpu->mmio_fragments[0];
 
-	memcpy(vcpu->run->mmio.data, frag->data, frag->len);
+	memcpy(vcpu->run->mmio.data, frag->data, min(8u, frag->len));
 	return X86EMUL_CONTINUE;
 }
 
@@ -3832,18 +4115,11 @@ mmio:
 	bytes -= handled;
 	val += handled;
 
-	while (bytes) {
-		unsigned now = min(bytes, 8U);
-
-		frag = &vcpu->mmio_fragments[vcpu->mmio_nr_fragments++];
-		frag->gpa = gpa;
-		frag->data = val;
-		frag->len = now;
-
-		gpa += now;
-		val += now;
-		bytes -= now;
-	}
+	WARN_ON(vcpu->mmio_nr_fragments >= KVM_MAX_MMIO_FRAGMENTS);
+	frag = &vcpu->mmio_fragments[vcpu->mmio_nr_fragments++];
+	frag->gpa = gpa;
+	frag->data = val;
+	frag->len = bytes;
 	return X86EMUL_CONTINUE;
 }
 
@@ -3890,7 +4166,7 @@ int emulator_read_write(struct x86_emulate_ctxt *ctxt, unsigned long addr,
 	vcpu->mmio_needed = 1;
 	vcpu->mmio_cur_fragment = 0;
 
-	vcpu->run->mmio.len = vcpu->mmio_fragments[0].len;
+	vcpu->run->mmio.len = min(8u, vcpu->mmio_fragments[0].len);
 	vcpu->run->mmio.is_write = vcpu->mmio_is_write = ops->write;
 	vcpu->run->exit_reason = KVM_EXIT_MMIO;
 	vcpu->run->mmio.phys_addr = gpa;
@@ -4280,7 +4556,12 @@ static int emulator_get_msr(struct x86_emulate_ctxt *ctxt,
 static int emulator_set_msr(struct x86_emulate_ctxt *ctxt,
 			    u32 msr_index, u64 data)
 {
-	return kvm_set_msr(emul_to_vcpu(ctxt), msr_index, data);
+	struct msr_data msr;
+
+	msr.data = data;
+	msr.index = msr_index;
+	msr.host_initiated = false;
+	return kvm_set_msr(emul_to_vcpu(ctxt), &msr);
 }
 
 static int emulator_read_pmc(struct x86_emulate_ctxt *ctxt,
@@ -4502,7 +4783,7 @@ static bool reexecute_instruction(struct kvm_vcpu *vcpu, gva_t gva)
 	 * instruction -> ...
 	 */
 	pfn = gfn_to_pfn(vcpu->kvm, gpa_to_gfn(gpa));
-	if (!is_error_pfn(pfn)) {
+	if (!is_error_noslot_pfn(pfn)) {
 		kvm_release_pfn_clean(pfn);
 		return true;
 	}
@@ -4888,6 +5169,50 @@ static void kvm_set_mmio_spte_mask(void)
 	kvm_mmu_set_mmio_spte_mask(mask);
 }
 
+#ifdef CONFIG_X86_64
+static void pvclock_gtod_update_fn(struct work_struct *work)
+{
+	struct kvm *kvm;
+
+	struct kvm_vcpu *vcpu;
+	int i;
+
+	raw_spin_lock(&kvm_lock);
+	list_for_each_entry(kvm, &vm_list, vm_list)
+		kvm_for_each_vcpu(i, vcpu, kvm)
+			set_bit(KVM_REQ_MASTERCLOCK_UPDATE, &vcpu->requests);
+	atomic_set(&kvm_guest_has_master_clock, 0);
+	raw_spin_unlock(&kvm_lock);
+}
+
+static DECLARE_WORK(pvclock_gtod_work, pvclock_gtod_update_fn);
+
+/*
+ * Notification about pvclock gtod data update.
+ */
+static int pvclock_gtod_notify(struct notifier_block *nb, unsigned long unused,
+			       void *priv)
+{
+	struct pvclock_gtod_data *gtod = &pvclock_gtod_data;
+	struct timekeeper *tk = priv;
+
+	update_pvclock_gtod(tk);
+
+	/* disable master clock if host does not trust, or does not
+	 * use, TSC clocksource
+	 */
+	if (gtod->clock.vclock_mode != VCLOCK_TSC &&
+	    atomic_read(&kvm_guest_has_master_clock) != 0)
+		queue_work(system_long_wq, &pvclock_gtod_work);
+
+	return 0;
+}
+
+static struct notifier_block pvclock_gtod_notifier = {
+	.notifier_call = pvclock_gtod_notify,
+};
+#endif
+
 int kvm_arch_init(void *opaque)
 {
 	int r;
@@ -4910,9 +5235,16 @@ int kvm_arch_init(void *opaque)
 		goto out;
 	}
 
+	r = -ENOMEM;
+	shared_msrs = alloc_percpu(struct kvm_shared_msrs);
+	if (!shared_msrs) {
+		printk(KERN_ERR "kvm: failed to allocate percpu kvm_shared_msrs\n");
+		goto out;
+	}
+
 	r = kvm_mmu_module_init();
 	if (r)
-		goto out;
+		goto out_free_percpu;
 
 	kvm_set_mmio_spte_mask();
 	kvm_init_msr_list();
@@ -4929,8 +5261,14 @@ int kvm_arch_init(void *opaque)
 		host_xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
 
 	kvm_lapic_init();
+#ifdef CONFIG_X86_64
+	pvclock_gtod_register_notifier(&pvclock_gtod_notifier);
+#endif
+
 	return 0;
 
+out_free_percpu:
+	free_percpu(shared_msrs);
 out:
 	return r;
 }
@@ -4943,8 +5281,12 @@ void kvm_arch_exit(void)
 		cpufreq_unregister_notifier(&kvmclock_cpufreq_notifier_block,
 					    CPUFREQ_TRANSITION_NOTIFIER);
 	unregister_hotcpu_notifier(&kvmclock_cpu_notifier_block);
+#ifdef CONFIG_X86_64
+	pvclock_gtod_unregister_notifier(&pvclock_gtod_notifier);
+#endif
 	kvm_x86_ops = NULL;
 	kvm_mmu_module_exit();
+	free_percpu(shared_msrs);
 }
 
 int kvm_emulate_halt(struct kvm_vcpu *vcpu)
@@ -5066,7 +5408,7 @@ out:
 }
 EXPORT_SYMBOL_GPL(kvm_emulate_hypercall);
 
-int emulator_fix_hypercall(struct x86_emulate_ctxt *ctxt)
+static int emulator_fix_hypercall(struct x86_emulate_ctxt *ctxt)
 {
 	struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt);
 	char instruction[3];
@@ -5242,6 +5584,29 @@ static void process_nmi(struct kvm_vcpu *vcpu)
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 }
 
+static void kvm_gen_update_masterclock(struct kvm *kvm)
+{
+#ifdef CONFIG_X86_64
+	int i;
+	struct kvm_vcpu *vcpu;
+	struct kvm_arch *ka = &kvm->arch;
+
+	spin_lock(&ka->pvclock_gtod_sync_lock);
+	kvm_make_mclock_inprogress_request(kvm);
+	/* no guest entries from this point */
+	pvclock_update_vm_gtod_copy(kvm);
+
+	kvm_for_each_vcpu(i, vcpu, kvm)
+		set_bit(KVM_REQ_CLOCK_UPDATE, &vcpu->requests);
+
+	/* guest entries allowed */
+	kvm_for_each_vcpu(i, vcpu, kvm)
+		clear_bit(KVM_REQ_MCLOCK_INPROGRESS, &vcpu->requests);
+
+	spin_unlock(&ka->pvclock_gtod_sync_lock);
+#endif
+}
+
 static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 {
 	int r;
@@ -5254,6 +5619,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 			kvm_mmu_unload(vcpu);
 		if (kvm_check_request(KVM_REQ_MIGRATE_TIMER, vcpu))
 			__kvm_migrate_timers(vcpu);
+		if (kvm_check_request(KVM_REQ_MASTERCLOCK_UPDATE, vcpu))
+			kvm_gen_update_masterclock(vcpu->kvm);
 		if (kvm_check_request(KVM_REQ_CLOCK_UPDATE, vcpu)) {
 			r = kvm_guest_time_update(vcpu);
 			if (unlikely(r))
@@ -5369,7 +5736,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 	if (hw_breakpoint_active())
 		hw_breakpoint_restore();
 
-	vcpu->arch.last_guest_tsc = kvm_x86_ops->read_l1_tsc(vcpu);
+	vcpu->arch.last_guest_tsc = kvm_x86_ops->read_l1_tsc(vcpu,
+							   native_read_tsc());
 
 	vcpu->mode = OUTSIDE_GUEST_MODE;
 	smp_wmb();
@@ -5426,7 +5794,7 @@ static int __vcpu_run(struct kvm_vcpu *vcpu)
 		pr_debug("vcpu %d received sipi with vector # %x\n",
 			 vcpu->vcpu_id, vcpu->arch.sipi_vector);
 		kvm_lapic_reset(vcpu);
-		r = kvm_arch_vcpu_reset(vcpu);
+		r = kvm_vcpu_reset(vcpu);
 		if (r)
 			return r;
 		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
@@ -5522,28 +5890,44 @@ static int complete_emulated_pio(struct kvm_vcpu *vcpu)
  *
  * read:
  *   for each fragment
- *     write gpa, len
- *     exit
- *     copy data
+ *     for each mmio piece in the fragment
+ *       write gpa, len
+ *       exit
+ *       copy data
  *   execute insn
  *
  * write:
  *   for each fragment
- *      write gpa, len
- *      copy data
- *      exit
+ *     for each mmio piece in the fragment
+ *       write gpa, len
+ *       copy data
+ *       exit
  */
 static int complete_emulated_mmio(struct kvm_vcpu *vcpu)
 {
 	struct kvm_run *run = vcpu->run;
 	struct kvm_mmio_fragment *frag;
+	unsigned len;
 
 	BUG_ON(!vcpu->mmio_needed);
 
 	/* Complete previous fragment */
-	frag = &vcpu->mmio_fragments[vcpu->mmio_cur_fragment++];
+	frag = &vcpu->mmio_fragments[vcpu->mmio_cur_fragment];
+	len = min(8u, frag->len);
 	if (!vcpu->mmio_is_write)
-		memcpy(frag->data, run->mmio.data, frag->len);
+		memcpy(frag->data, run->mmio.data, len);
+
+	if (frag->len <= 8) {
+		/* Switch to the next fragment. */
+		frag++;
+		vcpu->mmio_cur_fragment++;
+	} else {
+		/* Go forward to the next mmio piece. */
+		frag->data += len;
+		frag->gpa += len;
+		frag->len -= len;
+	}
+
 	if (vcpu->mmio_cur_fragment == vcpu->mmio_nr_fragments) {
 		vcpu->mmio_needed = 0;
 		if (vcpu->mmio_is_write)
@@ -5551,13 +5935,12 @@ static int complete_emulated_mmio(struct kvm_vcpu *vcpu)
 		vcpu->mmio_read_completed = 1;
 		return complete_emulated_io(vcpu);
 	}
-	/* Initiate next fragment */
-	++frag;
+
 	run->exit_reason = KVM_EXIT_MMIO;
 	run->mmio.phys_addr = frag->gpa;
 	if (vcpu->mmio_is_write)
-		memcpy(run->mmio.data, frag->data, frag->len);
-	run->mmio.len = frag->len;
+		memcpy(run->mmio.data, frag->data, min(8u, frag->len));
+	run->mmio.len = min(8u, frag->len);
 	run->mmio.is_write = vcpu->mmio_is_write;
 	vcpu->arch.complete_userspace_io = complete_emulated_mmio;
 	return 0;
@@ -5773,6 +6156,9 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
 
+	if (!guest_cpuid_has_xsave(vcpu) && (sregs->cr4 & X86_CR4_OSXSAVE))
+		return -EINVAL;
+
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
 	kvm_x86_ops->set_idt(vcpu, &dt);
@@ -6036,7 +6422,7 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 	r = vcpu_load(vcpu);
 	if (r)
 		return r;
-	r = kvm_arch_vcpu_reset(vcpu);
+	r = kvm_vcpu_reset(vcpu);
 	if (r == 0)
 		r = kvm_mmu_setup(vcpu);
 	vcpu_put(vcpu);
@@ -6044,6 +6430,23 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 	return r;
 }
 
+int kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
+{
+	int r;
+	struct msr_data msr;
+
+	r = vcpu_load(vcpu);
+	if (r)
+		return r;
+	msr.data = 0x0;
+	msr.index = MSR_IA32_TSC;
+	msr.host_initiated = true;
+	kvm_write_tsc(vcpu, &msr);
+	vcpu_put(vcpu);
+
+	return r;
+}
+
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
 	int r;
@@ -6058,7 +6461,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 	kvm_x86_ops->vcpu_free(vcpu);
 }
 
-int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
+static int kvm_vcpu_reset(struct kvm_vcpu *vcpu)
 {
 	atomic_set(&vcpu->arch.nmi_queued, 0);
 	vcpu->arch.nmi_pending = 0;
@@ -6081,6 +6484,10 @@ int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
 
 	kvm_pmu_reset(vcpu);
 
+	memset(vcpu->arch.regs, 0, sizeof(vcpu->arch.regs));
+	vcpu->arch.regs_avail = ~0;
+	vcpu->arch.regs_dirty = ~0;
+
 	return kvm_x86_ops->vcpu_reset(vcpu);
 }
 
@@ -6157,6 +6564,8 @@ int kvm_arch_hardware_enable(void *garbage)
 			kvm_for_each_vcpu(i, vcpu, kvm) {
 				vcpu->arch.tsc_offset_adjustment += delta_cyc;
 				vcpu->arch.last_host_tsc = local_tsc;
+				set_bit(KVM_REQ_MASTERCLOCK_UPDATE,
+					&vcpu->requests);
 			}
 
 			/*
@@ -6247,10 +6656,17 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 	if (!zalloc_cpumask_var(&vcpu->arch.wbinvd_dirty_mask, GFP_KERNEL))
 		goto fail_free_mce_banks;
 
+	r = fx_init(vcpu);
+	if (r)
+		goto fail_free_wbinvd_dirty_mask;
+
+	vcpu->arch.ia32_tsc_adjust_msr = 0x0;
 	kvm_async_pf_hash_reset(vcpu);
 	kvm_pmu_init(vcpu);
 
 	return 0;
+fail_free_wbinvd_dirty_mask:
+	free_cpumask_var(vcpu->arch.wbinvd_dirty_mask);
 fail_free_mce_banks:
 	kfree(vcpu->arch.mce_banks);
 fail_free_lapic:
@@ -6294,6 +6710,9 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
 	raw_spin_lock_init(&kvm->arch.tsc_write_lock);
 	mutex_init(&kvm->arch.apic_map_lock);
+	spin_lock_init(&kvm->arch.pvclock_gtod_sync_lock);
+
+	pvclock_update_vm_gtod_copy(kvm);
 
 	return 0;
 }