KVM: arm/arm64: Update the physical timer interrupt level
authorJintack Lim <jintack@cs.columbia.edu>
Fri, 3 Feb 2017 15:20:04 +0000 (10:20 -0500)
committerMarc Zyngier <marc.zyngier@arm.com>
Wed, 8 Feb 2017 15:13:35 +0000 (15:13 +0000)
Now that we maintain the EL1 physical timer register states of VMs,
update the physical timer interrupt level along with the virtual one.

Signed-off-by: Jintack Lim <jintack@cs.columbia.edu>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
virt/kvm/arm/arch_timer.c

index dbd0af19d27ec8bdc1b6f89bf24127f92755de7b..7f9a664199911ae27e7e188eb51309e5e6a9f9af 100644 (file)
@@ -186,6 +186,7 @@ static int kvm_timer_update_state(struct kvm_vcpu *vcpu)
 {
        struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
        struct arch_timer_context *vtimer = vcpu_vtimer(vcpu);
+       struct arch_timer_context *ptimer = vcpu_ptimer(vcpu);
 
        /*
         * If userspace modified the timer registers via SET_ONE_REG before
@@ -199,6 +200,9 @@ static int kvm_timer_update_state(struct kvm_vcpu *vcpu)
        if (kvm_timer_should_fire(vtimer) != vtimer->irq.level)
                kvm_timer_update_irq(vcpu, !vtimer->irq.level, vtimer);
 
+       if (kvm_timer_should_fire(ptimer) != ptimer->irq.level)
+               kvm_timer_update_irq(vcpu, !ptimer->irq.level, ptimer);
+
        return 0;
 }