svm: Add interrupt injection via AVIC

This patch introduces a new mechanism to inject interrupt using AVIC.
Since VINTR is not supported when enable AVIC, we need to inject
interrupt via APIC backing page instead.

This patch also adds support for AVIC doorbell, which is used by
KVM to signal a running vcpu to check IRR for injected interrupts.

Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 192cad25d3219dbac9b31abe75aaec08c887b92a..95d9b5c4c4319ebf6e8f388939904052e93b4481 100644 (file)
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -71,6 +71,8 @@ MODULE_DEVICE_TABLE(x86cpu, svm_cpu_id);
 #define SVM_FEATURE_DECODE_ASSIST  (1 <<  7)
 #define SVM_FEATURE_PAUSE_FILTER   (1 << 10)
 
+#define SVM_AVIC_DOORBELL      0xc001011b
+
 #define NESTED_EXIT_HOST       0       /* Exit handled on host level */
 #define NESTED_EXIT_DONE       1       /* Exit caused nested vmexit  */
 #define NESTED_EXIT_CONTINUE   2       /* Further checks needed      */
@@ -293,6 +295,17 @@ static inline void avic_update_vapic_bar(struct vcpu_svm *svm, u64 data)
        mark_dirty(svm->vmcb, VMCB_AVIC);
 }
 
+static inline bool avic_vcpu_is_running(struct kvm_vcpu *vcpu)
+{
+       struct vcpu_svm *svm = to_svm(vcpu);
+       u64 *entry = svm->avic_physical_id_cache;
+
+       if (!entry)
+               return false;
+
+       return (READ_ONCE(*entry) & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK);
+}
+
 static void recalc_intercepts(struct vcpu_svm *svm)
 {
        struct vmcb_control_area *c, *h;
@@ -2866,10 +2879,11 @@ static int clgi_interception(struct vcpu_svm *svm)
        disable_gif(svm);
 
        /* After a CLGI no interrupts should come */
-       svm_clear_vintr(svm);
-       svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
-
-       mark_dirty(svm->vmcb, VMCB_INTR);
+       if (!kvm_vcpu_apicv_active(&svm->vcpu)) {
+               svm_clear_vintr(svm);
+               svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
+               mark_dirty(svm->vmcb, VMCB_INTR);
+       }
 
        return 1;
 }
@@ -3763,6 +3777,7 @@ static inline void svm_inject_irq(struct vcpu_svm *svm, int irq)
 {
        struct vmcb_control_area *control;
 
+       /* The following fields are ignored when AVIC is enabled */
        control = &svm->vmcb->control;
        control->int_vector = irq;
        control->int_ctl &= ~V_INTR_PRIO_MASK;
@@ -3841,6 +3856,18 @@ static void svm_sync_pir_to_irr(struct kvm_vcpu *vcpu)
        return;
 }
 
+static void svm_deliver_avic_intr(struct kvm_vcpu *vcpu, int vec)
+{
+       kvm_lapic_set_irr(vec, vcpu->arch.apic);
+       smp_mb__after_atomic();
+
+       if (avic_vcpu_is_running(vcpu))
+               wrmsrl(SVM_AVIC_DOORBELL,
+                      __default_cpu_present_to_apicid(vcpu->cpu));
+       else
+               kvm_vcpu_wake_up(vcpu);
+}
+
 static int svm_nmi_allowed(struct kvm_vcpu *vcpu)
 {
        struct vcpu_svm *svm = to_svm(vcpu);
@@ -3895,6 +3922,9 @@ static void enable_irq_window(struct kvm_vcpu *vcpu)
 {
        struct vcpu_svm *svm = to_svm(vcpu);
 
+       if (kvm_vcpu_apicv_active(vcpu))
+               return;
+
        /*
         * In case GIF=0 we can't rely on the CPU to tell us when GIF becomes
         * 1, because that's a separate STGI/VMRUN intercept.  The next time we
@@ -4638,6 +4668,7 @@ static struct kvm_x86_ops svm_x86_ops = {
        .sched_in = svm_sched_in,
 
        .pmu_ops = &amd_pmu_ops,
+       .deliver_posted_interrupt = svm_deliver_avic_intr,
 };
 
 static int __init svm_init(void)