Merge tag 'kvmarm-fixes-6.4-3' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD

KVM/arm64 fixes for 6.4, take #3

- Fix the reported address of a watchpoint forwarded to userspace

- Fix the freeing of the root of stage-2 page tables

- Stop creating spurious PMU events to perform detection of the
  default PMU and use the existing PMU list instead.

Author: bonzini

arch/arm64/kvm/hyp/include/hyp/switch.h

@@ -412,17 +412,21 @@ static bool kvm_hyp_handle_cp15_32(struct kvm_vcpu *vcpu, u64 *exit_code)
 	return false;
 }
 
-static bool kvm_hyp_handle_iabt_low(struct kvm_vcpu *vcpu, u64 *exit_code)
+static bool kvm_hyp_handle_memory_fault(struct kvm_vcpu *vcpu, u64 *exit_code)
 {
 	if (!__populate_fault_info(vcpu))
 		return true;
 
 	return false;
 }
+static bool kvm_hyp_handle_iabt_low(struct kvm_vcpu *vcpu, u64 *exit_code)
+	__alias(kvm_hyp_handle_memory_fault);
+static bool kvm_hyp_handle_watchpt_low(struct kvm_vcpu *vcpu, u64 *exit_code)
+	__alias(kvm_hyp_handle_memory_fault);
 
 static bool kvm_hyp_handle_dabt_low(struct kvm_vcpu *vcpu, u64 *exit_code)
 {
-	if (!__populate_fault_info(vcpu))
+	if (kvm_hyp_handle_memory_fault(vcpu, exit_code))
 		return true;
 
 	if (static_branch_unlikely(&vgic_v2_cpuif_trap)) {

arch/arm64/kvm/hyp/nvhe/switch.c

@@ -186,6 +186,7 @@ static const exit_handler_fn hyp_exit_handlers[] = {
 	[ESR_ELx_EC_FP_ASIMD]		= kvm_hyp_handle_fpsimd,
 	[ESR_ELx_EC_IABT_LOW]		= kvm_hyp_handle_iabt_low,
 	[ESR_ELx_EC_DABT_LOW]		= kvm_hyp_handle_dabt_low,
+	[ESR_ELx_EC_WATCHPT_LOW]	= kvm_hyp_handle_watchpt_low,
 	[ESR_ELx_EC_PAC]		= kvm_hyp_handle_ptrauth,
 };
 
@@ -196,6 +197,7 @@ static const exit_handler_fn pvm_exit_handlers[] = {
 	[ESR_ELx_EC_FP_ASIMD]		= kvm_hyp_handle_fpsimd,
 	[ESR_ELx_EC_IABT_LOW]		= kvm_hyp_handle_iabt_low,
 	[ESR_ELx_EC_DABT_LOW]		= kvm_hyp_handle_dabt_low,
+	[ESR_ELx_EC_WATCHPT_LOW]	= kvm_hyp_handle_watchpt_low,
 	[ESR_ELx_EC_PAC]		= kvm_hyp_handle_ptrauth,
 };
 

arch/arm64/kvm/hyp/pgtable.c

@@ -1332,4 +1332,7 @@ void kvm_pgtable_stage2_free_removed(struct kvm_pgtable_mm_ops *mm_ops, void *pg
 	};
 
 	WARN_ON(__kvm_pgtable_walk(&data, mm_ops, ptep, level + 1));
+
+	WARN_ON(mm_ops->page_count(pgtable) != 1);
+	mm_ops->put_page(pgtable);
 }

arch/arm64/kvm/hyp/vhe/switch.c

@@ -110,6 +110,7 @@ static const exit_handler_fn hyp_exit_handlers[] = {
 	[ESR_ELx_EC_FP_ASIMD]		= kvm_hyp_handle_fpsimd,
 	[ESR_ELx_EC_IABT_LOW]		= kvm_hyp_handle_iabt_low,
 	[ESR_ELx_EC_DABT_LOW]		= kvm_hyp_handle_dabt_low,
+	[ESR_ELx_EC_WATCHPT_LOW]	= kvm_hyp_handle_watchpt_low,
 	[ESR_ELx_EC_PAC]		= kvm_hyp_handle_ptrauth,
 };
 

arch/arm64/kvm/pmu-emul.c

@@ -694,45 +694,23 @@ void kvm_host_pmu_init(struct arm_pmu *pmu)
 
 static struct arm_pmu *kvm_pmu_probe_armpmu(void)
 {
-	struct perf_event_attr attr = { };
-	struct perf_event *event;
-	struct arm_pmu *pmu = NULL;
-
-	/*
-	 * Create a dummy event that only counts user cycles. As we'll never
-	 * leave this function with the event being live, it will never
-	 * count anything. But it allows us to probe some of the PMU
-	 * details. Yes, this is terrible.
-	 */
-	attr.type = PERF_TYPE_RAW;
-	attr.size = sizeof(attr);
-	attr.pinned = 1;
-	attr.disabled = 0;
-	attr.exclude_user = 0;
-	attr.exclude_kernel = 1;
-	attr.exclude_hv = 1;
-	attr.exclude_host = 1;
-	attr.config = ARMV8_PMUV3_PERFCTR_CPU_CYCLES;
-	attr.sample_period = GENMASK(63, 0);
+	struct arm_pmu *tmp, *pmu = NULL;
+	struct arm_pmu_entry *entry;
+	int cpu;
 
-	event = perf_event_create_kernel_counter(&attr, -1, current,
-						 kvm_pmu_perf_overflow, &attr);
+	mutex_lock(&arm_pmus_lock);
 
-	if (IS_ERR(event)) {
-		pr_err_once("kvm: pmu event creation failed %ld\n",
-			    PTR_ERR(event));
-		return NULL;
-	}
+	cpu = smp_processor_id();
+	list_for_each_entry(entry, &arm_pmus, entry) {
+		tmp = entry->arm_pmu;
 
-	if (event->pmu) {
-		pmu = to_arm_pmu(event->pmu);
-		if (pmu->pmuver == ID_AA64DFR0_EL1_PMUVer_NI ||
-		    pmu->pmuver == ID_AA64DFR0_EL1_PMUVer_IMP_DEF)
-			pmu = NULL;
+		if (cpumask_test_cpu(cpu, &tmp->supported_cpus)) {
+			pmu = tmp;
+			break;
+		}
 	}
 
-	perf_event_disable(event);
-	perf_event_release_kernel(event);
+	mutex_unlock(&arm_pmus_lock);
 
 	return pmu;
 }
@@ -912,7 +890,17 @@ int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 		return -EBUSY;
 
 	if (!kvm->arch.arm_pmu) {
-		/* No PMU set, get the default one */
+		/*
+		 * No PMU set, get the default one.
+		 *
+		 * The observant among you will notice that the supported_cpus
+		 * mask does not get updated for the default PMU even though it
+		 * is quite possible the selected instance supports only a
+		 * subset of cores in the system. This is intentional, and
+		 * upholds the preexisting behavior on heterogeneous systems
+		 * where vCPUs can be scheduled on any core but the guest
+		 * counters could stop working.
+		 */
 		kvm->arch.arm_pmu = kvm_pmu_probe_armpmu();
 		if (!kvm->arch.arm_pmu)
 			return -ENODEV;