Merge branch kvm-arm64/m1 into kvmarm-master/next

Rework the KVM GIC and timer to cope with lesser HW such as
the Apple M1 SoC.

* kvm-arm64/m1:
  irqchip/apple-aic: Advertise some level of vGICv3 compatibility
  KVM: arm64: timer: Add support for SW-based deactivation
  KVM: arm64: timer: Refactor IRQ configuration
  KVM: arm64: vgic: Implement SW-driven deactivation
  KVM: arm64: vgic: move irq->get_input_level into an ops structure
  KVM: arm64: vgic: Let an interrupt controller advertise lack of HW deactivation
  KVM: arm64: vgic: Be tolerant to the lack of maintenance interrupt masking
  KVM: arm64: Handle physical FIQ as an IRQ while running a guest
  irqchip/gic: Split vGIC probing information from the GIC code

Author: Marc Zyngier

arch/arm64/kvm/arch_timer.c

@@ -9,6 +9,7 @@
 #include <linux/kvm_host.h>
 #include <linux/interrupt.h>
 #include <linux/irq.h>
+#include <linux/irqdomain.h>
 #include <linux/uaccess.h>
 
 #include <clocksource/arm_arch_timer.h>
@@ -973,36 +974,154 @@ static int kvm_timer_dying_cpu(unsigned int cpu)
 	return 0;
 }
 
-int kvm_timer_hyp_init(bool has_gic)
+static int timer_irq_set_vcpu_affinity(struct irq_data *d, void *vcpu)
 {
-	struct arch_timer_kvm_info *info;
-	int err;
+	if (vcpu)
+		irqd_set_forwarded_to_vcpu(d);
+	else
+		irqd_clr_forwarded_to_vcpu(d);
 
-	info = arch_timer_get_kvm_info();
-	timecounter = &info->timecounter;
+	return 0;
+}
 
-	if (!timecounter->cc) {
-		kvm_err("kvm_arch_timer: uninitialized timecounter\n");
-		return -ENODEV;
+static int timer_irq_set_irqchip_state(struct irq_data *d,
+				       enum irqchip_irq_state which, bool val)
+{
+	if (which != IRQCHIP_STATE_ACTIVE || !irqd_is_forwarded_to_vcpu(d))
+		return irq_chip_set_parent_state(d, which, val);
+
+	if (val)
+		irq_chip_mask_parent(d);
+	else
+		irq_chip_unmask_parent(d);
+
+	return 0;
+}
+
+static void timer_irq_eoi(struct irq_data *d)
+{
+	if (!irqd_is_forwarded_to_vcpu(d))
+		irq_chip_eoi_parent(d);
+}
+
+static void timer_irq_ack(struct irq_data *d)
+{
+	d = d->parent_data;
+	if (d->chip->irq_ack)
+		d->chip->irq_ack(d);
+}
+
+static struct irq_chip timer_chip = {
+	.name			= "KVM",
+	.irq_ack		= timer_irq_ack,
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= timer_irq_eoi,
+	.irq_set_type		= irq_chip_set_type_parent,
+	.irq_set_vcpu_affinity	= timer_irq_set_vcpu_affinity,
+	.irq_set_irqchip_state	= timer_irq_set_irqchip_state,
+};
+
+static int timer_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				  unsigned int nr_irqs, void *arg)
+{
+	irq_hw_number_t hwirq = (uintptr_t)arg;
+
+	return irq_domain_set_hwirq_and_chip(domain, virq, hwirq,
+					     &timer_chip, NULL);
+}
+
+static void timer_irq_domain_free(struct irq_domain *domain, unsigned int virq,
+				  unsigned int nr_irqs)
+{
+}
+
+static const struct irq_domain_ops timer_domain_ops = {
+	.alloc	= timer_irq_domain_alloc,
+	.free	= timer_irq_domain_free,
+};
+
+static struct irq_ops arch_timer_irq_ops = {
+	.get_input_level = kvm_arch_timer_get_input_level,
+};
+
+static void kvm_irq_fixup_flags(unsigned int virq, u32 *flags)
+{
+	*flags = irq_get_trigger_type(virq);
+	if (*flags != IRQF_TRIGGER_HIGH && *flags != IRQF_TRIGGER_LOW) {
+		kvm_err("Invalid trigger for timer IRQ%d, assuming level low\n",
+			virq);
+		*flags = IRQF_TRIGGER_LOW;
 	}
+}
 
-	/* First, do the virtual EL1 timer irq */
+static int kvm_irq_init(struct arch_timer_kvm_info *info)
+{
+	struct irq_domain *domain = NULL;
 
 	if (info->virtual_irq <= 0) {
 		kvm_err("kvm_arch_timer: invalid virtual timer IRQ: %d\n",
 			info->virtual_irq);
 		return -ENODEV;
 	}
+
 	host_vtimer_irq = info->virtual_irq;
+	kvm_irq_fixup_flags(host_vtimer_irq, &host_vtimer_irq_flags);
+
+	if (kvm_vgic_global_state.no_hw_deactivation) {
+		struct fwnode_handle *fwnode;
+		struct irq_data *data;
+
+		fwnode = irq_domain_alloc_named_fwnode("kvm-timer");
+		if (!fwnode)
+			return -ENOMEM;
+
+		/* Assume both vtimer and ptimer in the same parent */
+		data = irq_get_irq_data(host_vtimer_irq);
+		domain = irq_domain_create_hierarchy(data->domain, 0,
+						     NR_KVM_TIMERS, fwnode,
+						     &timer_domain_ops, NULL);
+		if (!domain) {
+			irq_domain_free_fwnode(fwnode);
+			return -ENOMEM;
+		}
+
+		arch_timer_irq_ops.flags |= VGIC_IRQ_SW_RESAMPLE;
+		WARN_ON(irq_domain_push_irq(domain, host_vtimer_irq,
+					    (void *)TIMER_VTIMER));
+	}
 
-	host_vtimer_irq_flags = irq_get_trigger_type(host_vtimer_irq);
-	if (host_vtimer_irq_flags != IRQF_TRIGGER_HIGH &&
-	    host_vtimer_irq_flags != IRQF_TRIGGER_LOW) {
-		kvm_err("Invalid trigger for vtimer IRQ%d, assuming level low\n",
-			host_vtimer_irq);
-		host_vtimer_irq_flags = IRQF_TRIGGER_LOW;
+	if (info->physical_irq > 0) {
+		host_ptimer_irq = info->physical_irq;
+		kvm_irq_fixup_flags(host_ptimer_irq, &host_ptimer_irq_flags);
+
+		if (domain)
+			WARN_ON(irq_domain_push_irq(domain, host_ptimer_irq,
+						    (void *)TIMER_PTIMER));
 	}
 
+	return 0;
+}
+
+int kvm_timer_hyp_init(bool has_gic)
+{
+	struct arch_timer_kvm_info *info;
+	int err;
+
+	info = arch_timer_get_kvm_info();
+	timecounter = &info->timecounter;
+
+	if (!timecounter->cc) {
+		kvm_err("kvm_arch_timer: uninitialized timecounter\n");
+		return -ENODEV;
+	}
+
+	err = kvm_irq_init(info);
+	if (err)
+		return err;
+
+	/* First, do the virtual EL1 timer irq */
+
 	err = request_percpu_irq(host_vtimer_irq, kvm_arch_timer_handler,
 				 "kvm guest vtimer", kvm_get_running_vcpus());
 	if (err) {
@@ -1027,15 +1146,6 @@ int kvm_timer_hyp_init(bool has_gic)
 	/* Now let's do the physical EL1 timer irq */
 
 	if (info->physical_irq > 0) {
-		host_ptimer_irq = info->physical_irq;
-		host_ptimer_irq_flags = irq_get_trigger_type(host_ptimer_irq);
-		if (host_ptimer_irq_flags != IRQF_TRIGGER_HIGH &&
-		    host_ptimer_irq_flags != IRQF_TRIGGER_LOW) {
-			kvm_err("Invalid trigger for ptimer IRQ%d, assuming level low\n",
-				host_ptimer_irq);
-			host_ptimer_irq_flags = IRQF_TRIGGER_LOW;
-		}
-
 		err = request_percpu_irq(host_ptimer_irq, kvm_arch_timer_handler,
 					 "kvm guest ptimer", kvm_get_running_vcpus());
 		if (err) {
@@ -1143,15 +1253,15 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 	ret = kvm_vgic_map_phys_irq(vcpu,
 				    map.direct_vtimer->host_timer_irq,
 				    map.direct_vtimer->irq.irq,
-				    kvm_arch_timer_get_input_level);
+				    &arch_timer_irq_ops);
 	if (ret)
 		return ret;
 
 	if (map.direct_ptimer) {
 		ret = kvm_vgic_map_phys_irq(vcpu,
 					    map.direct_ptimer->host_timer_irq,
 					    map.direct_ptimer->irq.irq,
-					    kvm_arch_timer_get_input_level);
+					    &arch_timer_irq_ops);
 	}
 
 	if (ret)

arch/arm64/kvm/hyp/hyp-entry.S

@@ -76,6 +76,7 @@ el1_trap:
 	b	__guest_exit
 
 el1_irq:
+el1_fiq:
 	get_vcpu_ptr	x1, x0
 	mov	x0, #ARM_EXCEPTION_IRQ
 	b	__guest_exit
@@ -131,7 +132,6 @@ SYM_CODE_END(\label)
 	invalid_vector	el2t_error_invalid
 	invalid_vector	el2h_irq_invalid
 	invalid_vector	el2h_fiq_invalid
-	invalid_vector	el1_fiq_invalid
 
 	.ltorg
 
@@ -179,12 +179,12 @@ SYM_CODE_START(__kvm_hyp_vector)
 
 	valid_vect	el1_sync		// Synchronous 64-bit EL1
 	valid_vect	el1_irq			// IRQ 64-bit EL1
-	invalid_vect	el1_fiq_invalid		// FIQ 64-bit EL1
+	valid_vect	el1_fiq			// FIQ 64-bit EL1
 	valid_vect	el1_error		// Error 64-bit EL1
 
 	valid_vect	el1_sync		// Synchronous 32-bit EL1
 	valid_vect	el1_irq			// IRQ 32-bit EL1
-	invalid_vect	el1_fiq_invalid		// FIQ 32-bit EL1
+	valid_vect	el1_fiq			// FIQ 32-bit EL1
 	valid_vect	el1_error		// Error 32-bit EL1
 SYM_CODE_END(__kvm_hyp_vector)
 

arch/arm64/kvm/vgic/vgic-init.c

@@ -482,6 +482,16 @@ static irqreturn_t vgic_maintenance_handler(int irq, void *data)
 	return IRQ_HANDLED;
 }
 
+static struct gic_kvm_info *gic_kvm_info;
+
+void __init vgic_set_kvm_info(const struct gic_kvm_info *info)
+{
+	BUG_ON(gic_kvm_info != NULL);
+	gic_kvm_info = kmalloc(sizeof(*info), GFP_KERNEL);
+	if (gic_kvm_info)
+		*gic_kvm_info = *info;
+}
+
 /**
  * kvm_vgic_init_cpu_hardware - initialize the GIC VE hardware
  *
@@ -509,18 +519,29 @@ void kvm_vgic_init_cpu_hardware(void)
  */
 int kvm_vgic_hyp_init(void)
 {
-	const struct gic_kvm_info *gic_kvm_info;
+	bool has_mask;
 	int ret;
 
-	gic_kvm_info = gic_get_kvm_info();
 	if (!gic_kvm_info)
 		return -ENODEV;
 
-	if (!gic_kvm_info->maint_irq) {
+	has_mask = !gic_kvm_info->no_maint_irq_mask;
+
+	if (has_mask && !gic_kvm_info->maint_irq) {
 		kvm_err("No vgic maintenance irq\n");
 		return -ENXIO;
 	}
 
+	/*
+	 * If we get one of these oddball non-GICs, taint the kernel,
+	 * as we have no idea of how they *really* behave.
+	 */
+	if (gic_kvm_info->no_hw_deactivation) {
+		kvm_info("Non-architectural vgic, tainting kernel\n");
+		add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK);
+		kvm_vgic_global_state.no_hw_deactivation = true;
+	}
+
 	switch (gic_kvm_info->type) {
 	case GIC_V2:
 		ret = vgic_v2_probe(gic_kvm_info);
@@ -536,10 +557,17 @@ int kvm_vgic_hyp_init(void)
 		ret = -ENODEV;
 	}
 
+	kvm_vgic_global_state.maint_irq = gic_kvm_info->maint_irq;
+
+	kfree(gic_kvm_info);
+	gic_kvm_info = NULL;
+
 	if (ret)
 		return ret;
 
-	kvm_vgic_global_state.maint_irq = gic_kvm_info->maint_irq;
+	if (!has_mask)
+		return 0;
+
 	ret = request_percpu_irq(kvm_vgic_global_state.maint_irq,
 				 vgic_maintenance_handler,
 				 "vgic", kvm_get_running_vcpus());

arch/arm64/kvm/vgic/vgic-v2.c

@@ -108,11 +108,22 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 		 * If this causes us to lower the level, we have to also clear
 		 * the physical active state, since we will otherwise never be
 		 * told when the interrupt becomes asserted again.
+		 *
+		 * Another case is when the interrupt requires a helping hand
+		 * on deactivation (no HW deactivation, for example).
 		 */
-		if (vgic_irq_is_mapped_level(irq) && (val & GICH_LR_PENDING_BIT)) {
-			irq->line_level = vgic_get_phys_line_level(irq);
+		if (vgic_irq_is_mapped_level(irq)) {
+			bool resample = false;
+
+			if (val & GICH_LR_PENDING_BIT) {
+				irq->line_level = vgic_get_phys_line_level(irq);
+				resample = !irq->line_level;
+			} else if (vgic_irq_needs_resampling(irq) &&
+				   !(irq->active || irq->pending_latch)) {
+				resample = true;
+			}
 
-			if (!irq->line_level)
+			if (resample)
 				vgic_irq_set_phys_active(irq, false);
 		}
 
@@ -152,7 +163,7 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 	if (irq->group)
 		val |= GICH_LR_GROUP1;
 
-	if (irq->hw) {
+	if (irq->hw && !vgic_irq_needs_resampling(irq)) {
 		val |= GICH_LR_HW;
 		val |= irq->hwintid << GICH_LR_PHYSID_CPUID_SHIFT;
 		/*

arch/arm64/kvm/vgic/vgic-v3.c

@@ -101,11 +101,22 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 		 * If this causes us to lower the level, we have to also clear
 		 * the physical active state, since we will otherwise never be
 		 * told when the interrupt becomes asserted again.
+		 *
+		 * Another case is when the interrupt requires a helping hand
+		 * on deactivation (no HW deactivation, for example).
 		 */
-		if (vgic_irq_is_mapped_level(irq) && (val & ICH_LR_PENDING_BIT)) {
-			irq->line_level = vgic_get_phys_line_level(irq);
+		if (vgic_irq_is_mapped_level(irq)) {
+			bool resample = false;
+
+			if (val & ICH_LR_PENDING_BIT) {
+				irq->line_level = vgic_get_phys_line_level(irq);
+				resample = !irq->line_level;
+			} else if (vgic_irq_needs_resampling(irq) &&
+				   !(irq->active || irq->pending_latch)) {
+				resample = true;
+			}
 
-			if (!irq->line_level)
+			if (resample)
 				vgic_irq_set_phys_active(irq, false);
 		}
 
@@ -136,7 +147,7 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 		}
 	}
 
-	if (irq->hw) {
+	if (irq->hw && !vgic_irq_needs_resampling(irq)) {
 		val |= ICH_LR_HW;
 		val |= ((u64)irq->hwintid) << ICH_LR_PHYS_ID_SHIFT;
 		/*