From c3d7a5283382d5238ef23cd900ad3ab38b473753 Mon Sep 17 00:00:00 2001
From: Zachary Amsden <zamsden@redhat.com>
Date: Thu, 19 Aug 2010 22:07:20 -1000
Subject: [PATCH] --- yaml --- r: 215763 b: refs/heads/master c:
 f38e098ff3a315bb74abbb4a35cba11bbea8e2fa h: refs/heads/master i:   215761:
 3dfc70a5070a398862c1538bf4cf04b46fecef29   215759:
 607a9cabd144acf4de156bc345a6f10c9a5bb67a v: v3

---
 [refs]                                |  2 +-
 trunk/arch/x86/include/asm/kvm_host.h |  3 +++
 trunk/arch/x86/kvm/x86.c              | 31 ++++++++++++++++++++++++++-
 3 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/[refs] b/[refs]
index aeab718c07ba..810bd5538acc 100644
--- a/[refs]
+++ b/[refs]
@@ -1,2 +1,2 @@
 ---
-refs/heads/master: 99e3e30aee1a326a98bf3a5f47b8622219c685f3
+refs/heads/master: f38e098ff3a315bb74abbb4a35cba11bbea8e2fa
diff --git a/trunk/arch/x86/include/asm/kvm_host.h b/trunk/arch/x86/include/asm/kvm_host.h
index a215153f1ff6..57b4394491ec 100644
--- a/trunk/arch/x86/include/asm/kvm_host.h
+++ b/trunk/arch/x86/include/asm/kvm_host.h
@@ -396,6 +396,9 @@ struct kvm_arch {
 	unsigned long irq_sources_bitmap;
 	s64 kvmclock_offset;
 	spinlock_t tsc_write_lock;
+	u64 last_tsc_nsec;
+	u64 last_tsc_offset;
+	u64 last_tsc_write;
 
 	struct kvm_xen_hvm_config xen_hvm_config;
 
diff --git a/trunk/arch/x86/kvm/x86.c b/trunk/arch/x86/kvm/x86.c
index 886132b6ef14..e7da14c317e6 100644
--- a/trunk/arch/x86/kvm/x86.c
+++ b/trunk/arch/x86/kvm/x86.c
@@ -898,11 +898,40 @@ static DEFINE_PER_CPU(unsigned long, cpu_tsc_khz);
 void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 {
 	struct kvm *kvm = vcpu->kvm;
-	u64 offset;
+	u64 offset, ns, elapsed;
 	unsigned long flags;
+	struct timespec ts;
 
 	spin_lock_irqsave(&kvm->arch.tsc_write_lock, flags);
 	offset = data - native_read_tsc();
+	ktime_get_ts(&ts);
+	monotonic_to_bootbased(&ts);
+	ns = timespec_to_ns(&ts);
+	elapsed = ns - kvm->arch.last_tsc_nsec;
+
+	/*
+	 * Special case: identical write to TSC within 5 seconds of
+	 * another CPU is interpreted as an attempt to synchronize
+	 * (the 5 seconds is to accomodate host load / swapping).
+	 *
+	 * In that case, for a reliable TSC, we can match TSC offsets,
+	 * or make a best guest using kernel_ns value.
+	 */
+	if (data == kvm->arch.last_tsc_write && elapsed < 5ULL * NSEC_PER_SEC) {
+		if (!check_tsc_unstable()) {
+			offset = kvm->arch.last_tsc_offset;
+			pr_debug("kvm: matched tsc offset for %llu\n", data);
+		} else {
+			u64 tsc_delta = elapsed * __get_cpu_var(cpu_tsc_khz);
+			tsc_delta = tsc_delta / USEC_PER_SEC;
+			offset += tsc_delta;
+			pr_debug("kvm: adjusted tsc offset by %llu\n", tsc_delta);
+		}
+		ns = kvm->arch.last_tsc_nsec;
+	}
+	kvm->arch.last_tsc_nsec = ns;
+	kvm->arch.last_tsc_write = data;
+	kvm->arch.last_tsc_offset = offset;
 	kvm_x86_ops->write_tsc_offset(vcpu, offset);
 	spin_unlock_irqrestore(&kvm->arch.tsc_write_lock, flags);