1 files changed, 153 insertions, 51 deletions

diff --git a/arch/arm64/kvm/vgic/vgic-v4.c b/arch/arm64/kvm/vgic/vgic-v4.c
index 27ac833e5ec7..09c3e9eb23f8 100644
--- a/arch/arm64/kvm/vgic/vgic-v4.c
+++ b/arch/arm64/kvm/vgic/vgic-v4.c
@@ -90,7 +90,15 @@ static irqreturn_t vgic_v4_doorbell_handler(int irq, void *info)
 	    !irqd_irq_disabled(&irq_to_desc(irq)->irq_data))
 		disable_irq_nosync(irq);
 
+	/*
+	 * The v4.1 doorbell can fire concurrently with the vPE being
+	 * made non-resident. Ensure we only update pending_last
+	 * *after* the non-residency sequence has completed.
+	 */
+	raw_spin_lock(&vcpu->arch.vgic_cpu.vgic_v3.its_vpe.vpe_lock);
 	vcpu->arch.vgic_cpu.vgic_v3.its_vpe.pending_last = true;
+	raw_spin_unlock(&vcpu->arch.vgic_cpu.vgic_v3.its_vpe.vpe_lock);
+
 	kvm_make_request(KVM_REQ_IRQ_PENDING, vcpu);
 	kvm_vcpu_kick(vcpu);
 
@@ -115,7 +123,7 @@ static void vgic_v4_enable_vsgis(struct kvm_vcpu *vcpu)
 	 * IRQ. The SGI code will do its magic.
 	 */
 	for (i = 0; i < VGIC_NR_SGIS; i++) {
-		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, i);
+		struct vgic_irq *irq = vgic_get_vcpu_irq(vcpu, i);
 		struct irq_desc *desc;
 		unsigned long flags;
 		int ret;
@@ -152,9 +160,10 @@ static void vgic_v4_disable_vsgis(struct kvm_vcpu *vcpu)
 	int i;
 
 	for (i = 0; i < VGIC_NR_SGIS; i++) {
-		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, i);
+		struct vgic_irq *irq = vgic_get_vcpu_irq(vcpu, i);
 		struct irq_desc *desc;
 		unsigned long flags;
+		bool pending;
 		int ret;
 
 		raw_spin_lock_irqsave(&irq->irq_lock, flags);
@@ -165,9 +174,11 @@ static void vgic_v4_disable_vsgis(struct kvm_vcpu *vcpu)
 		irq->hw = false;
 		ret = irq_get_irqchip_state(irq->host_irq,
 					    IRQCHIP_STATE_PENDING,
-					    &irq->pending_latch);
+					    &pending);
 		WARN_ON(ret);
 
+		irq->pending_latch = pending;
+
 		desc = irq_to_desc(irq->host_irq);
 		irq_domain_deactivate_irq(irq_desc_get_irq_data(desc));
 	unlock:
@@ -176,12 +187,13 @@ static void vgic_v4_disable_vsgis(struct kvm_vcpu *vcpu)
 	}
 }
 
-/* Must be called with the kvm lock held */
 void vgic_v4_configure_vsgis(struct kvm *kvm)
 {
 	struct vgic_dist *dist = &kvm->arch.vgic;
 	struct kvm_vcpu *vcpu;
-	int i;
+	unsigned long i;
+
+	lockdep_assert_held(&kvm->arch.config_lock);
 
 	kvm_arm_halt_guest(kvm);
 
@@ -195,20 +207,46 @@ void vgic_v4_configure_vsgis(struct kvm *kvm)
 	kvm_arm_resume_guest(kvm);
 }
 
+/*
+ * Must be called with GICv4.1 and the vPE unmapped, which
+ * indicates the invalidation of any VPT caches associated
+ * with the vPE, thus we can get the VLPI state by peeking
+ * at the VPT.
+ */
+void vgic_v4_get_vlpi_state(struct vgic_irq *irq, bool *val)
+{
+	struct its_vpe *vpe = &irq->target_vcpu->arch.vgic_cpu.vgic_v3.its_vpe;
+	int mask = BIT(irq->intid % BITS_PER_BYTE);
+	void *va;
+	u8 *ptr;
+
+	va = page_address(vpe->vpt_page);
+	ptr = va + irq->intid / BITS_PER_BYTE;
+
+	*val = !!(*ptr & mask);
+}
+
+int vgic_v4_request_vpe_irq(struct kvm_vcpu *vcpu, int irq)
+{
+	return request_irq(irq, vgic_v4_doorbell_handler, 0, "vcpu", vcpu);
+}
+
 /**
  * vgic_v4_init - Initialize the GICv4 data structures
  * @kvm:	Pointer to the VM being initialized
  *
  * We may be called each time a vITS is created, or when the
- * vgic is initialized. This relies on kvm->lock to be
- * held. In both cases, the number of vcpus should now be
- * fixed.
+ * vgic is initialized. In both cases, the number of vcpus
+ * should now be fixed.
  */
 int vgic_v4_init(struct kvm *kvm)
 {
 	struct vgic_dist *dist = &kvm->arch.vgic;
 	struct kvm_vcpu *vcpu;
-	int i, nr_vcpus, ret;
+	int nr_vcpus, ret;
+	unsigned long i;
+
+	lockdep_assert_held(&kvm->arch.config_lock);
 
 	if (!kvm_vgic_global_state.has_gicv4)
 		return 0; /* Nothing to see here... move along. */
@@ -219,7 +257,7 @@ int vgic_v4_init(struct kvm *kvm)
 	nr_vcpus = atomic_read(&kvm->online_vcpus);
 
 	dist->its_vm.vpes = kcalloc(nr_vcpus, sizeof(*dist->its_vm.vpes),
-				    GFP_KERNEL);
+				    GFP_KERNEL_ACCOUNT);
 	if (!dist->its_vm.vpes)
 		return -ENOMEM;
 
@@ -255,8 +293,7 @@ int vgic_v4_init(struct kvm *kvm)
 			irq_flags &= ~IRQ_NOAUTOEN;
 		irq_set_status_flags(irq, irq_flags);
 
-		ret = request_irq(irq, vgic_v4_doorbell_handler,
-				  0, "vcpu", vcpu);
+		ret = vgic_v4_request_vpe_irq(vcpu, irq);
 		if (ret) {
 			kvm_err("failed to allocate vcpu IRQ%d\n", irq);
 			/*
@@ -277,14 +314,14 @@ int vgic_v4_init(struct kvm *kvm)
 /**
  * vgic_v4_teardown - Free the GICv4 data structures
  * @kvm:	Pointer to the VM being destroyed
- *
- * Relies on kvm->lock to be held.
  */
 void vgic_v4_teardown(struct kvm *kvm)
 {
 	struct its_vm *its_vm = &kvm->arch.vgic.its_vm;
 	int i;
 
+	lockdep_assert_held(&kvm->arch.config_lock);
+
 	if (!its_vm->vpes)
 		return;
 
@@ -302,14 +339,30 @@ void vgic_v4_teardown(struct kvm *kvm)
 	its_vm->vpes = NULL;
 }
 
-int vgic_v4_put(struct kvm_vcpu *vcpu, bool need_db)
+static inline bool vgic_v4_want_doorbell(struct kvm_vcpu *vcpu)
+{
+	if (vcpu_get_flag(vcpu, IN_WFI))
+		return true;
+
+	if (likely(!vcpu_has_nv(vcpu)))
+		return false;
+
+	/*
+	 * GICv4 hardware is only ever used for the L1. Mark the vPE (i.e. the
+	 * L1 context) nonresident and request a doorbell to kick us out of the
+	 * L2 when an IRQ becomes pending.
+	 */
+	return vcpu_get_flag(vcpu, IN_NESTED_ERET);
+}
+
+int vgic_v4_put(struct kvm_vcpu *vcpu)
 {
 	struct its_vpe *vpe = &vcpu->arch.vgic_cpu.vgic_v3.its_vpe;
 
-	if (!vgic_supports_direct_msis(vcpu->kvm) || !vpe->resident)
+	if (!vgic_supports_direct_irqs(vcpu->kvm) || !vpe->resident)
 		return 0;
 
-	return its_make_vpe_non_resident(vpe, need_db);
+	return its_make_vpe_non_resident(vpe, vgic_v4_want_doorbell(vcpu));
 }
 
 int vgic_v4_load(struct kvm_vcpu *vcpu)
@@ -317,7 +370,10 @@ int vgic_v4_load(struct kvm_vcpu *vcpu)
 	struct its_vpe *vpe = &vcpu->arch.vgic_cpu.vgic_v3.its_vpe;
 	int err;
 
-	if (!vgic_supports_direct_msis(vcpu->kvm) || vpe->resident)
+	if (!vgic_supports_direct_irqs(vcpu->kvm) || vpe->resident)
+		return 0;
+
+	if (vcpu_get_flag(vcpu, IN_WFI))
 		return 0;
 
 	/*
@@ -345,6 +401,18 @@ int vgic_v4_load(struct kvm_vcpu *vcpu)
 	return err;
 }
 
+void vgic_v4_commit(struct kvm_vcpu *vcpu)
+{
+	struct its_vpe *vpe = &vcpu->arch.vgic_cpu.vgic_v3.its_vpe;
+
+	/*
+	 * No need to wait for the vPE to be ready across a shallow guest
+	 * exit, as only a vcpu_put will invalidate it.
+	 */
+	if (!vpe->ready)
+		its_commit_vpe(vpe);
+}
+
 static struct vgic_its *vgic_get_its(struct kvm *kvm,
 				     struct kvm_kernel_irq_routing_entry *irq_entry)
 {
@@ -365,7 +433,8 @@ int kvm_vgic_v4_set_forwarding(struct kvm *kvm, int virq,
 	struct vgic_its *its;
 	struct vgic_irq *irq;
 	struct its_vlpi_map map;
-	int ret;
+	unsigned long flags;
+	int ret = 0;
 
 	if (!vgic_supports_direct_msis(kvm))
 		return 0;
@@ -378,13 +447,24 @@ int kvm_vgic_v4_set_forwarding(struct kvm *kvm, int virq,
 	if (IS_ERR(its))
 		return 0;
 
-	mutex_lock(&its->its_lock);
+	guard(mutex)(&its->its_lock);
 
-	/* Perform the actual DevID/EventID -> LPI translation. */
-	ret = vgic_its_resolve_lpi(kvm, its, irq_entry->msi.devid,
-				   irq_entry->msi.data, &irq);
-	if (ret)
-		goto out;
+	/*
+	 * Perform the actual DevID/EventID -> LPI translation.
+	 *
+	 * Silently exit if translation fails as the guest (or userspace!) has
+	 * managed to do something stupid. Emulated LPI injection will still
+	 * work if the guest figures itself out at a later time.
+	 */
+	if (vgic_its_resolve_lpi(kvm, its, irq_entry->msi.devid,
+				 irq_entry->msi.data, &irq))
+		return 0;
+
+	raw_spin_lock_irqsave(&irq->irq_lock, flags);
+
+	/* Silently exit if the vLPI is already mapped */
+	if (irq->hw)
+		goto out_unlock_irq;
 
 	/*
 	 * Emit the mapping request. If it fails, the ITS probably
@@ -404,50 +484,72 @@ int kvm_vgic_v4_set_forwarding(struct kvm *kvm, int virq,
 
 	ret = its_map_vlpi(virq, &map);
 	if (ret)
-		goto out;
+		goto out_unlock_irq;
 
 	irq->hw		= true;
 	irq->host_irq	= virq;
 	atomic_inc(&map.vpe->vlpi_count);
 
-out:
-	mutex_unlock(&its->its_lock);
+	/* Transfer pending state */
+	if (!irq->pending_latch)
+		goto out_unlock_irq;
+
+	ret = irq_set_irqchip_state(irq->host_irq, IRQCHIP_STATE_PENDING,
+				    irq->pending_latch);
+	WARN_RATELIMIT(ret, "IRQ %d", irq->host_irq);
+
+	/*
+	 * Clear pending_latch and communicate this state
+	 * change via vgic_queue_irq_unlock.
+	 */
+	irq->pending_latch = false;
+	vgic_queue_irq_unlock(kvm, irq, flags);
+	return ret;
+
+out_unlock_irq:
+	raw_spin_unlock_irqrestore(&irq->irq_lock, flags);
 	return ret;
 }
 
-int kvm_vgic_v4_unset_forwarding(struct kvm *kvm, int virq,
-				 struct kvm_kernel_irq_routing_entry *irq_entry)
+static struct vgic_irq *__vgic_host_irq_get_vlpi(struct kvm *kvm, int host_irq)
 {
-	struct vgic_its *its;
 	struct vgic_irq *irq;
-	int ret;
+	unsigned long idx;
 
-	if (!vgic_supports_direct_msis(kvm))
-		return 0;
+	guard(rcu)();
+	xa_for_each(&kvm->arch.vgic.lpi_xa, idx, irq) {
+		if (!irq->hw || irq->host_irq != host_irq)
+			continue;
 
-	/*
-	 * Get the ITS, and escape early on error (not a valid
-	 * doorbell for any of our vITSs).
-	 */
-	its = vgic_get_its(kvm, irq_entry);
-	if (IS_ERR(its))
-		return 0;
+		if (!vgic_try_get_irq_ref(irq))
+			return NULL;
 
-	mutex_lock(&its->its_lock);
+		return irq;
+	}
 
-	ret = vgic_its_resolve_lpi(kvm, its, irq_entry->msi.devid,
-				   irq_entry->msi.data, &irq);
-	if (ret)
-		goto out;
+	return NULL;
+}
 
-	WARN_ON(!(irq->hw && irq->host_irq == virq));
+void kvm_vgic_v4_unset_forwarding(struct kvm *kvm, int host_irq)
+{
+	struct vgic_irq *irq;
+	unsigned long flags;
+
+	if (!vgic_supports_direct_msis(kvm))
+		return;
+
+	irq = __vgic_host_irq_get_vlpi(kvm, host_irq);
+	if (!irq)
+		return;
+
+	raw_spin_lock_irqsave(&irq->irq_lock, flags);
+	WARN_ON(irq->hw && irq->host_irq != host_irq);
 	if (irq->hw) {
 		atomic_dec(&irq->target_vcpu->arch.vgic_cpu.vgic_v3.its_vpe.vlpi_count);
 		irq->hw = false;
-		ret = its_unmap_vlpi(virq);
+		its_unmap_vlpi(host_irq);
 	}
 
-out:
-	mutex_unlock(&its->its_lock);
-	return ret;
+	raw_spin_unlock_irqrestore(&irq->irq_lock, flags);
+	vgic_put_irq(kvm, irq);
 }