KVM: ppc: Handle guest idle by emulating MSR[WE] writes

This reduces host CPU usage when the guest is idle. However, the guest must
set MSR[WE] in its idle loop, which Linux did not do until 2.6.26.

Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Jerone Young <jyoung5@us.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>

4 files changed, 24 insertions, 3 deletions

diff --git a/arch/powerpc/kvm/booke_guest.c b/arch/powerpc/kvm/booke_guest.c
index 6d9884a..b3db6f4 100644
--- a/arch/powerpc/kvm/booke_guest.c
+++ b/arch/powerpc/kvm/booke_guest.c
@@ -49,6 +49,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "inst_emu",   VCPU_STAT(emulated_inst_exits) },
 	{ "dec",        VCPU_STAT(dec_exits) },
 	{ "ext_intr",   VCPU_STAT(ext_intr_exits) },
+	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
 	{ NULL }
 };
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index bad40bd..777e0f3 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -36,13 +36,12 @@ gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
 
 int kvm_cpu_has_interrupt(struct kvm_vcpu *v)
 {
-	/* XXX implement me */
-	return 0;
+	return !!(v->arch.pending_exceptions);
 }
 
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
 {
-	return 1;
+	return !(v->arch.msr & MSR_WE);
 }
 
 
@@ -214,6 +213,11 @@ static void kvmppc_decrementer_func(unsigned long data)
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu *)data;
 
 	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_DECREMENTER);
+
+	if (waitqueue_active(&vcpu->wq)) {
+		wake_up_interruptible(&vcpu->wq);
+		vcpu->stat.halt_wakeup++;
+	}
 }
 
 int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
@@ -339,6 +343,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
@@ -363,12 +369,20 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
+
 	return r;
 }
 
 int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
 {
 	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_EXTERNAL);
+
+	if (waitqueue_active(&vcpu->wq)) {
+		wake_up_interruptible(&vcpu->wq);
+		vcpu->stat.halt_wakeup++;
+	}
+
 	return 0;
 }
 
diff --git a/include/asm-powerpc/kvm_host.h b/include/asm-powerpc/kvm_host.h
index 04ffbb8..81a69d7 100644
--- a/include/asm-powerpc/kvm_host.h
+++ b/include/asm-powerpc/kvm_host.h
@@ -59,6 +59,7 @@ struct kvm_vcpu_stat {
 	u32 emulated_inst_exits;
 	u32 dec_exits;
 	u32 ext_intr_exits;
+	u32 halt_wakeup;
 };
 
 struct tlbe {
diff --git a/include/asm-powerpc/kvm_ppc.h b/include/asm-powerpc/kvm_ppc.h
index 7ac8203..b35a7e3 100644
--- a/include/asm-powerpc/kvm_ppc.h
+++ b/include/asm-powerpc/kvm_ppc.h
@@ -77,12 +77,17 @@ static inline void kvmppc_clear_exception(struct kvm_vcpu *vcpu, int exception)
 	clear_bit(priority, &vcpu->arch.pending_exceptions);
 }
 
+/* Helper function for "full" MSR writes. No need to call this if only EE is
+ * changing. */
 static inline void kvmppc_set_msr(struct kvm_vcpu *vcpu, u32 new_msr)
 {
 	if ((new_msr & MSR_PR) != (vcpu->arch.msr & MSR_PR))
 		kvmppc_mmu_priv_switch(vcpu, new_msr & MSR_PR);
 
 	vcpu->arch.msr = new_msr;
+
+	if (vcpu->arch.msr & MSR_WE)
+		kvm_vcpu_block(vcpu);
 }
 
 #endif /* __POWERPC_KVM_PPC_H__ */