Adding upstream version 6.6.15.upstream/6.6.15

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

11 files changed, 5610 insertions, 0 deletions

diff --git a/tools/testing/selftests/kvm/aarch64/aarch32_id_regs.c b/tools/testing/selftests/kvm/aarch64/aarch32_id_regs.c
new file mode 100644
index 0000000000..b90580840b
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/aarch32_id_regs.c
@@ -0,0 +1,167 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * aarch32_id_regs - Test for ID register behavior on AArch64-only systems
+ *
+ * Copyright (c) 2022 Google LLC.
+ *
+ * Test that KVM handles the AArch64 views of the AArch32 ID registers as RAZ
+ * and WI from userspace.
+ */
+
+#include <stdint.h>
+
+#include "kvm_util.h"
+#include "processor.h"
+#include "test_util.h"
+#include <linux/bitfield.h>
+
+#define BAD_ID_REG_VAL	0x1badc0deul
+
+#define GUEST_ASSERT_REG_RAZ(reg)	GUEST_ASSERT_EQ(read_sysreg_s(reg), 0)
+
+static void guest_main(void)
+{
+	GUEST_ASSERT_REG_RAZ(SYS_ID_PFR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_PFR1_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_DFR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_AFR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR1_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR2_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR3_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR1_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR2_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR3_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR4_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR5_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR4_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_ISAR6_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_MVFR0_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_MVFR1_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_MVFR2_EL1);
+	GUEST_ASSERT_REG_RAZ(sys_reg(3, 0, 0, 3, 3));
+	GUEST_ASSERT_REG_RAZ(SYS_ID_PFR2_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_DFR1_EL1);
+	GUEST_ASSERT_REG_RAZ(SYS_ID_MMFR5_EL1);
+	GUEST_ASSERT_REG_RAZ(sys_reg(3, 0, 0, 3, 7));
+
+	GUEST_DONE();
+}
+
+static void test_guest_raz(struct kvm_vcpu *vcpu)
+{
+	struct ucall uc;
+
+	vcpu_run(vcpu);
+
+	switch (get_ucall(vcpu, &uc)) {
+	case UCALL_ABORT:
+		REPORT_GUEST_ASSERT(uc);
+		break;
+	case UCALL_DONE:
+		break;
+	default:
+		TEST_FAIL("Unexpected ucall: %lu", uc.cmd);
+	}
+}
+
+static uint64_t raz_wi_reg_ids[] = {
+	KVM_ARM64_SYS_REG(SYS_ID_PFR0_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_PFR1_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_DFR0_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR0_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR1_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR2_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR3_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR0_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR1_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR2_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR3_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR4_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR5_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR4_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_ISAR6_EL1),
+	KVM_ARM64_SYS_REG(SYS_MVFR0_EL1),
+	KVM_ARM64_SYS_REG(SYS_MVFR1_EL1),
+	KVM_ARM64_SYS_REG(SYS_MVFR2_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_PFR2_EL1),
+	KVM_ARM64_SYS_REG(SYS_ID_MMFR5_EL1),
+};
+
+static void test_user_raz_wi(struct kvm_vcpu *vcpu)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(raz_wi_reg_ids); i++) {
+		uint64_t reg_id = raz_wi_reg_ids[i];
+		uint64_t val;
+
+		vcpu_get_reg(vcpu, reg_id, &val);
+		TEST_ASSERT_EQ(val, 0);
+
+		/*
+		 * Expect the ioctl to succeed with no effect on the register
+		 * value.
+		 */
+		vcpu_set_reg(vcpu, reg_id, BAD_ID_REG_VAL);
+
+		vcpu_get_reg(vcpu, reg_id, &val);
+		TEST_ASSERT_EQ(val, 0);
+	}
+}
+
+static uint64_t raz_invariant_reg_ids[] = {
+	KVM_ARM64_SYS_REG(SYS_ID_AFR0_EL1),
+	KVM_ARM64_SYS_REG(sys_reg(3, 0, 0, 3, 3)),
+	KVM_ARM64_SYS_REG(SYS_ID_DFR1_EL1),
+	KVM_ARM64_SYS_REG(sys_reg(3, 0, 0, 3, 7)),
+};
+
+static void test_user_raz_invariant(struct kvm_vcpu *vcpu)
+{
+	int i, r;
+
+	for (i = 0; i < ARRAY_SIZE(raz_invariant_reg_ids); i++) {
+		uint64_t reg_id = raz_invariant_reg_ids[i];
+		uint64_t val;
+
+		vcpu_get_reg(vcpu, reg_id, &val);
+		TEST_ASSERT_EQ(val, 0);
+
+		r = __vcpu_set_reg(vcpu, reg_id, BAD_ID_REG_VAL);
+		TEST_ASSERT(r < 0 && errno == EINVAL,
+			    "unexpected KVM_SET_ONE_REG error: r=%d, errno=%d", r, errno);
+
+		vcpu_get_reg(vcpu, reg_id, &val);
+		TEST_ASSERT_EQ(val, 0);
+	}
+}
+
+
+
+static bool vcpu_aarch64_only(struct kvm_vcpu *vcpu)
+{
+	uint64_t val, el0;
+
+	vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(SYS_ID_AA64PFR0_EL1), &val);
+
+	el0 = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64PFR0_EL0), val);
+	return el0 == ID_AA64PFR0_ELx_64BIT_ONLY;
+}
+
+int main(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+
+	vm = vm_create_with_one_vcpu(&vcpu, guest_main);
+
+	TEST_REQUIRE(vcpu_aarch64_only(vcpu));
+
+	test_user_raz_wi(vcpu);
+	test_user_raz_invariant(vcpu);
+	test_guest_raz(vcpu);
+
+	kvm_vm_free(vm);
+}
diff --git a/tools/testing/selftests/kvm/aarch64/arch_timer.c b/tools/testing/selftests/kvm/aarch64/arch_timer.c
new file mode 100644
index 0000000000..274b8465b4
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/arch_timer.c
@@ -0,0 +1,480 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * arch_timer.c - Tests the aarch64 timer IRQ functionality
+ *
+ * The test validates both the virtual and physical timer IRQs using
+ * CVAL and TVAL registers. This consitutes the four stages in the test.
+ * The guest's main thread configures the timer interrupt for a stage
+ * and waits for it to fire, with a timeout equal to the timer period.
+ * It asserts that the timeout doesn't exceed the timer period.
+ *
+ * On the other hand, upon receipt of an interrupt, the guest's interrupt
+ * handler validates the interrupt by checking if the architectural state
+ * is in compliance with the specifications.
+ *
+ * The test provides command-line options to configure the timer's
+ * period (-p), number of vCPUs (-n), and iterations per stage (-i).
+ * To stress-test the timer stack even more, an option to migrate the
+ * vCPUs across pCPUs (-m), at a particular rate, is also provided.
+ *
+ * Copyright (c) 2021, Google LLC.
+ */
+#define _GNU_SOURCE
+
+#include <stdlib.h>
+#include <pthread.h>
+#include <linux/kvm.h>
+#include <linux/sizes.h>
+#include <linux/bitmap.h>
+#include <sys/sysinfo.h>
+
+#include "kvm_util.h"
+#include "processor.h"
+#include "delay.h"
+#include "arch_timer.h"
+#include "gic.h"
+#include "vgic.h"
+
+#define NR_VCPUS_DEF			4
+#define NR_TEST_ITERS_DEF		5
+#define TIMER_TEST_PERIOD_MS_DEF	10
+#define TIMER_TEST_ERR_MARGIN_US	100
+#define TIMER_TEST_MIGRATION_FREQ_MS	2
+
+struct test_args {
+	int nr_vcpus;
+	int nr_iter;
+	int timer_period_ms;
+	int migration_freq_ms;
+	struct kvm_arm_counter_offset offset;
+};
+
+static struct test_args test_args = {
+	.nr_vcpus = NR_VCPUS_DEF,
+	.nr_iter = NR_TEST_ITERS_DEF,
+	.timer_period_ms = TIMER_TEST_PERIOD_MS_DEF,
+	.migration_freq_ms = TIMER_TEST_MIGRATION_FREQ_MS,
+	.offset = { .reserved = 1 },
+};
+
+#define msecs_to_usecs(msec)		((msec) * 1000LL)
+
+#define GICD_BASE_GPA			0x8000000ULL
+#define GICR_BASE_GPA			0x80A0000ULL
+
+enum guest_stage {
+	GUEST_STAGE_VTIMER_CVAL = 1,
+	GUEST_STAGE_VTIMER_TVAL,
+	GUEST_STAGE_PTIMER_CVAL,
+	GUEST_STAGE_PTIMER_TVAL,
+	GUEST_STAGE_MAX,
+};
+
+/* Shared variables between host and guest */
+struct test_vcpu_shared_data {
+	int nr_iter;
+	enum guest_stage guest_stage;
+	uint64_t xcnt;
+};
+
+static struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
+static pthread_t pt_vcpu_run[KVM_MAX_VCPUS];
+static struct test_vcpu_shared_data vcpu_shared_data[KVM_MAX_VCPUS];
+
+static int vtimer_irq, ptimer_irq;
+
+static unsigned long *vcpu_done_map;
+static pthread_mutex_t vcpu_done_map_lock;
+
+static void
+guest_configure_timer_action(struct test_vcpu_shared_data *shared_data)
+{
+	switch (shared_data->guest_stage) {
+	case GUEST_STAGE_VTIMER_CVAL:
+		timer_set_next_cval_ms(VIRTUAL, test_args.timer_period_ms);
+		shared_data->xcnt = timer_get_cntct(VIRTUAL);
+		timer_set_ctl(VIRTUAL, CTL_ENABLE);
+		break;
+	case GUEST_STAGE_VTIMER_TVAL:
+		timer_set_next_tval_ms(VIRTUAL, test_args.timer_period_ms);
+		shared_data->xcnt = timer_get_cntct(VIRTUAL);
+		timer_set_ctl(VIRTUAL, CTL_ENABLE);
+		break;
+	case GUEST_STAGE_PTIMER_CVAL:
+		timer_set_next_cval_ms(PHYSICAL, test_args.timer_period_ms);
+		shared_data->xcnt = timer_get_cntct(PHYSICAL);
+		timer_set_ctl(PHYSICAL, CTL_ENABLE);
+		break;
+	case GUEST_STAGE_PTIMER_TVAL:
+		timer_set_next_tval_ms(PHYSICAL, test_args.timer_period_ms);
+		shared_data->xcnt = timer_get_cntct(PHYSICAL);
+		timer_set_ctl(PHYSICAL, CTL_ENABLE);
+		break;
+	default:
+		GUEST_ASSERT(0);
+	}
+}
+
+static void guest_validate_irq(unsigned int intid,
+				struct test_vcpu_shared_data *shared_data)
+{
+	enum guest_stage stage = shared_data->guest_stage;
+	uint64_t xcnt = 0, xcnt_diff_us, cval = 0;
+	unsigned long xctl = 0;
+	unsigned int timer_irq = 0;
+	unsigned int accessor;
+
+	if (intid == IAR_SPURIOUS)
+		return;
+
+	switch (stage) {
+	case GUEST_STAGE_VTIMER_CVAL:
+	case GUEST_STAGE_VTIMER_TVAL:
+		accessor = VIRTUAL;
+		timer_irq = vtimer_irq;
+		break;
+	case GUEST_STAGE_PTIMER_CVAL:
+	case GUEST_STAGE_PTIMER_TVAL:
+		accessor = PHYSICAL;
+		timer_irq = ptimer_irq;
+		break;
+	default:
+		GUEST_ASSERT(0);
+		return;
+	}
+
+	xctl = timer_get_ctl(accessor);
+	if ((xctl & CTL_IMASK) || !(xctl & CTL_ENABLE))
+		return;
+
+	timer_set_ctl(accessor, CTL_IMASK);
+	xcnt = timer_get_cntct(accessor);
+	cval = timer_get_cval(accessor);
+
+	xcnt_diff_us = cycles_to_usec(xcnt - shared_data->xcnt);
+
+	/* Make sure we are dealing with the correct timer IRQ */
+	GUEST_ASSERT_EQ(intid, timer_irq);
+
+	/* Basic 'timer condition met' check */
+	__GUEST_ASSERT(xcnt >= cval,
+		       "xcnt = 0x%llx, cval = 0x%llx, xcnt_diff_us = 0x%llx",
+		       xcnt, cval, xcnt_diff_us);
+	__GUEST_ASSERT(xctl & CTL_ISTATUS, "xcnt = 0x%llx", xcnt);
+
+	WRITE_ONCE(shared_data->nr_iter, shared_data->nr_iter + 1);
+}
+
+static void guest_irq_handler(struct ex_regs *regs)
+{
+	unsigned int intid = gic_get_and_ack_irq();
+	uint32_t cpu = guest_get_vcpuid();
+	struct test_vcpu_shared_data *shared_data = &vcpu_shared_data[cpu];
+
+	guest_validate_irq(intid, shared_data);
+
+	gic_set_eoi(intid);
+}
+
+static void guest_run_stage(struct test_vcpu_shared_data *shared_data,
+				enum guest_stage stage)
+{
+	uint32_t irq_iter, config_iter;
+
+	shared_data->guest_stage = stage;
+	shared_data->nr_iter = 0;
+
+	for (config_iter = 0; config_iter < test_args.nr_iter; config_iter++) {
+		/* Setup the next interrupt */
+		guest_configure_timer_action(shared_data);
+
+		/* Setup a timeout for the interrupt to arrive */
+		udelay(msecs_to_usecs(test_args.timer_period_ms) +
+			TIMER_TEST_ERR_MARGIN_US);
+
+		irq_iter = READ_ONCE(shared_data->nr_iter);
+		GUEST_ASSERT_EQ(config_iter + 1, irq_iter);
+	}
+}
+
+static void guest_code(void)
+{
+	uint32_t cpu = guest_get_vcpuid();
+	struct test_vcpu_shared_data *shared_data = &vcpu_shared_data[cpu];
+
+	local_irq_disable();
+
+	gic_init(GIC_V3, test_args.nr_vcpus,
+		(void *)GICD_BASE_GPA, (void *)GICR_BASE_GPA);
+
+	timer_set_ctl(VIRTUAL, CTL_IMASK);
+	timer_set_ctl(PHYSICAL, CTL_IMASK);
+
+	gic_irq_enable(vtimer_irq);
+	gic_irq_enable(ptimer_irq);
+	local_irq_enable();
+
+	guest_run_stage(shared_data, GUEST_STAGE_VTIMER_CVAL);
+	guest_run_stage(shared_data, GUEST_STAGE_VTIMER_TVAL);
+	guest_run_stage(shared_data, GUEST_STAGE_PTIMER_CVAL);
+	guest_run_stage(shared_data, GUEST_STAGE_PTIMER_TVAL);
+
+	GUEST_DONE();
+}
+
+static void *test_vcpu_run(void *arg)
+{
+	unsigned int vcpu_idx = (unsigned long)arg;
+	struct ucall uc;
+	struct kvm_vcpu *vcpu = vcpus[vcpu_idx];
+	struct kvm_vm *vm = vcpu->vm;
+	struct test_vcpu_shared_data *shared_data = &vcpu_shared_data[vcpu_idx];
+
+	vcpu_run(vcpu);
+
+	/* Currently, any exit from guest is an indication of completion */
+	pthread_mutex_lock(&vcpu_done_map_lock);
+	__set_bit(vcpu_idx, vcpu_done_map);
+	pthread_mutex_unlock(&vcpu_done_map_lock);
+
+	switch (get_ucall(vcpu, &uc)) {
+	case UCALL_SYNC:
+	case UCALL_DONE:
+		break;
+	case UCALL_ABORT:
+		sync_global_from_guest(vm, *shared_data);
+		fprintf(stderr, "Guest assert failed,  vcpu %u; stage; %u; iter: %u\n",
+			vcpu_idx, shared_data->guest_stage, shared_data->nr_iter);
+		REPORT_GUEST_ASSERT(uc);
+		break;
+	default:
+		TEST_FAIL("Unexpected guest exit\n");
+	}
+
+	return NULL;
+}
+
+static uint32_t test_get_pcpu(void)
+{
+	uint32_t pcpu;
+	unsigned int nproc_conf;
+	cpu_set_t online_cpuset;
+
+	nproc_conf = get_nprocs_conf();
+	sched_getaffinity(0, sizeof(cpu_set_t), &online_cpuset);
+
+	/* Randomly find an available pCPU to place a vCPU on */
+	do {
+		pcpu = rand() % nproc_conf;
+	} while (!CPU_ISSET(pcpu, &online_cpuset));
+
+	return pcpu;
+}
+
+static int test_migrate_vcpu(unsigned int vcpu_idx)
+{
+	int ret;
+	cpu_set_t cpuset;
+	uint32_t new_pcpu = test_get_pcpu();
+
+	CPU_ZERO(&cpuset);
+	CPU_SET(new_pcpu, &cpuset);
+
+	pr_debug("Migrating vCPU: %u to pCPU: %u\n", vcpu_idx, new_pcpu);
+
+	ret = pthread_setaffinity_np(pt_vcpu_run[vcpu_idx],
+				     sizeof(cpuset), &cpuset);
+
+	/* Allow the error where the vCPU thread is already finished */
+	TEST_ASSERT(ret == 0 || ret == ESRCH,
+		    "Failed to migrate the vCPU:%u to pCPU: %u; ret: %d\n",
+		    vcpu_idx, new_pcpu, ret);
+
+	return ret;
+}
+
+static void *test_vcpu_migration(void *arg)
+{
+	unsigned int i, n_done;
+	bool vcpu_done;
+
+	do {
+		usleep(msecs_to_usecs(test_args.migration_freq_ms));
+
+		for (n_done = 0, i = 0; i < test_args.nr_vcpus; i++) {
+			pthread_mutex_lock(&vcpu_done_map_lock);
+			vcpu_done = test_bit(i, vcpu_done_map);
+			pthread_mutex_unlock(&vcpu_done_map_lock);
+
+			if (vcpu_done) {
+				n_done++;
+				continue;
+			}
+
+			test_migrate_vcpu(i);
+		}
+	} while (test_args.nr_vcpus != n_done);
+
+	return NULL;
+}
+
+static void test_run(struct kvm_vm *vm)
+{
+	pthread_t pt_vcpu_migration;
+	unsigned int i;
+	int ret;
+
+	pthread_mutex_init(&vcpu_done_map_lock, NULL);
+	vcpu_done_map = bitmap_zalloc(test_args.nr_vcpus);
+	TEST_ASSERT(vcpu_done_map, "Failed to allocate vcpu done bitmap\n");
+
+	for (i = 0; i < (unsigned long)test_args.nr_vcpus; i++) {
+		ret = pthread_create(&pt_vcpu_run[i], NULL, test_vcpu_run,
+				     (void *)(unsigned long)i);
+		TEST_ASSERT(!ret, "Failed to create vCPU-%d pthread\n", i);
+	}
+
+	/* Spawn a thread to control the vCPU migrations */
+	if (test_args.migration_freq_ms) {
+		srand(time(NULL));
+
+		ret = pthread_create(&pt_vcpu_migration, NULL,
+					test_vcpu_migration, NULL);
+		TEST_ASSERT(!ret, "Failed to create the migration pthread\n");
+	}
+
+
+	for (i = 0; i < test_args.nr_vcpus; i++)
+		pthread_join(pt_vcpu_run[i], NULL);
+
+	if (test_args.migration_freq_ms)
+		pthread_join(pt_vcpu_migration, NULL);
+
+	bitmap_free(vcpu_done_map);
+}
+
+static void test_init_timer_irq(struct kvm_vm *vm)
+{
+	/* Timer initid should be same for all the vCPUs, so query only vCPU-0 */
+	vcpu_device_attr_get(vcpus[0], KVM_ARM_VCPU_TIMER_CTRL,
+			     KVM_ARM_VCPU_TIMER_IRQ_PTIMER, &ptimer_irq);
+	vcpu_device_attr_get(vcpus[0], KVM_ARM_VCPU_TIMER_CTRL,
+			     KVM_ARM_VCPU_TIMER_IRQ_VTIMER, &vtimer_irq);
+
+	sync_global_to_guest(vm, ptimer_irq);
+	sync_global_to_guest(vm, vtimer_irq);
+
+	pr_debug("ptimer_irq: %d; vtimer_irq: %d\n", ptimer_irq, vtimer_irq);
+}
+
+static int gic_fd;
+
+static struct kvm_vm *test_vm_create(void)
+{
+	struct kvm_vm *vm;
+	unsigned int i;
+	int nr_vcpus = test_args.nr_vcpus;
+
+	vm = vm_create_with_vcpus(nr_vcpus, guest_code, vcpus);
+
+	vm_init_descriptor_tables(vm);
+	vm_install_exception_handler(vm, VECTOR_IRQ_CURRENT, guest_irq_handler);
+
+	if (!test_args.offset.reserved) {
+		if (kvm_has_cap(KVM_CAP_COUNTER_OFFSET))
+			vm_ioctl(vm, KVM_ARM_SET_COUNTER_OFFSET, &test_args.offset);
+		else
+			TEST_FAIL("no support for global offset\n");
+	}
+
+	for (i = 0; i < nr_vcpus; i++)
+		vcpu_init_descriptor_tables(vcpus[i]);
+
+	test_init_timer_irq(vm);
+	gic_fd = vgic_v3_setup(vm, nr_vcpus, 64, GICD_BASE_GPA, GICR_BASE_GPA);
+	__TEST_REQUIRE(gic_fd >= 0, "Failed to create vgic-v3");
+
+	/* Make all the test's cmdline args visible to the guest */
+	sync_global_to_guest(vm, test_args);
+
+	return vm;
+}
+
+static void test_vm_cleanup(struct kvm_vm *vm)
+{
+	close(gic_fd);
+	kvm_vm_free(vm);
+}
+
+static void test_print_help(char *name)
+{
+	pr_info("Usage: %s [-h] [-n nr_vcpus] [-i iterations] [-p timer_period_ms]\n",
+		name);
+	pr_info("\t-n: Number of vCPUs to configure (default: %u; max: %u)\n",
+		NR_VCPUS_DEF, KVM_MAX_VCPUS);
+	pr_info("\t-i: Number of iterations per stage (default: %u)\n",
+		NR_TEST_ITERS_DEF);
+	pr_info("\t-p: Periodicity (in ms) of the guest timer (default: %u)\n",
+		TIMER_TEST_PERIOD_MS_DEF);
+	pr_info("\t-m: Frequency (in ms) of vCPUs to migrate to different pCPU. 0 to turn off (default: %u)\n",
+		TIMER_TEST_MIGRATION_FREQ_MS);
+	pr_info("\t-o: Counter offset (in counter cycles, default: 0)\n");
+	pr_info("\t-h: print this help screen\n");
+}
+
+static bool parse_args(int argc, char *argv[])
+{
+	int opt;
+
+	while ((opt = getopt(argc, argv, "hn:i:p:m:o:")) != -1) {
+		switch (opt) {
+		case 'n':
+			test_args.nr_vcpus = atoi_positive("Number of vCPUs", optarg);
+			if (test_args.nr_vcpus > KVM_MAX_VCPUS) {
+				pr_info("Max allowed vCPUs: %u\n",
+					KVM_MAX_VCPUS);
+				goto err;
+			}
+			break;
+		case 'i':
+			test_args.nr_iter = atoi_positive("Number of iterations", optarg);
+			break;
+		case 'p':
+			test_args.timer_period_ms = atoi_positive("Periodicity", optarg);
+			break;
+		case 'm':
+			test_args.migration_freq_ms = atoi_non_negative("Frequency", optarg);
+			break;
+		case 'o':
+			test_args.offset.counter_offset = strtol(optarg, NULL, 0);
+			test_args.offset.reserved = 0;
+			break;
+		case 'h':
+		default:
+			goto err;
+		}
+	}
+
+	return true;
+
+err:
+	test_print_help(argv[0]);
+	return false;
+}
+
+int main(int argc, char *argv[])
+{
+	struct kvm_vm *vm;
+
+	if (!parse_args(argc, argv))
+		exit(KSFT_SKIP);
+
+	__TEST_REQUIRE(!test_args.migration_freq_ms || get_nprocs() >= 2,
+		       "At least two physical CPUs needed for vCPU migration");
+
+	vm = test_vm_create();
+	test_run(vm);
+	test_vm_cleanup(vm);
+
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/debug-exceptions.c b/tools/testing/selftests/kvm/aarch64/debug-exceptions.c
new file mode 100644
index 0000000000..f5b6cb3a00
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/debug-exceptions.c
@@ -0,0 +1,607 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <test_util.h>
+#include <kvm_util.h>
+#include <processor.h>
+#include <linux/bitfield.h>
+
+#define MDSCR_KDE	(1 << 13)
+#define MDSCR_MDE	(1 << 15)
+#define MDSCR_SS	(1 << 0)
+
+#define DBGBCR_LEN8	(0xff << 5)
+#define DBGBCR_EXEC	(0x0 << 3)
+#define DBGBCR_EL1	(0x1 << 1)
+#define DBGBCR_E	(0x1 << 0)
+#define DBGBCR_LBN_SHIFT	16
+#define DBGBCR_BT_SHIFT		20
+#define DBGBCR_BT_ADDR_LINK_CTX	(0x1 << DBGBCR_BT_SHIFT)
+#define DBGBCR_BT_CTX_LINK	(0x3 << DBGBCR_BT_SHIFT)
+
+#define DBGWCR_LEN8	(0xff << 5)
+#define DBGWCR_RD	(0x1 << 3)
+#define DBGWCR_WR	(0x2 << 3)
+#define DBGWCR_EL1	(0x1 << 1)
+#define DBGWCR_E	(0x1 << 0)
+#define DBGWCR_LBN_SHIFT	16
+#define DBGWCR_WT_SHIFT		20
+#define DBGWCR_WT_LINK		(0x1 << DBGWCR_WT_SHIFT)
+
+#define SPSR_D		(1 << 9)
+#define SPSR_SS		(1 << 21)
+
+extern unsigned char sw_bp, sw_bp2, hw_bp, hw_bp2, bp_svc, bp_brk, hw_wp, ss_start, hw_bp_ctx;
+extern unsigned char iter_ss_begin, iter_ss_end;
+static volatile uint64_t sw_bp_addr, hw_bp_addr;
+static volatile uint64_t wp_addr, wp_data_addr;
+static volatile uint64_t svc_addr;
+static volatile uint64_t ss_addr[4], ss_idx;
+#define  PC(v)  ((uint64_t)&(v))
+
+#define GEN_DEBUG_WRITE_REG(reg_name)			\
+static void write_##reg_name(int num, uint64_t val)	\
+{							\
+	switch (num) {					\
+	case 0:						\
+		write_sysreg(val, reg_name##0_el1);	\
+		break;					\
+	case 1:						\
+		write_sysreg(val, reg_name##1_el1);	\
+		break;					\
+	case 2:						\
+		write_sysreg(val, reg_name##2_el1);	\
+		break;					\
+	case 3:						\
+		write_sysreg(val, reg_name##3_el1);	\
+		break;					\
+	case 4:						\
+		write_sysreg(val, reg_name##4_el1);	\
+		break;					\
+	case 5:						\
+		write_sysreg(val, reg_name##5_el1);	\
+		break;					\
+	case 6:						\
+		write_sysreg(val, reg_name##6_el1);	\
+		break;					\
+	case 7:						\
+		write_sysreg(val, reg_name##7_el1);	\
+		break;					\
+	case 8:						\
+		write_sysreg(val, reg_name##8_el1);	\
+		break;					\
+	case 9:						\
+		write_sysreg(val, reg_name##9_el1);	\
+		break;					\
+	case 10:					\
+		write_sysreg(val, reg_name##10_el1);	\
+		break;					\
+	case 11:					\
+		write_sysreg(val, reg_name##11_el1);	\
+		break;					\
+	case 12:					\
+		write_sysreg(val, reg_name##12_el1);	\
+		break;					\
+	case 13:					\
+		write_sysreg(val, reg_name##13_el1);	\
+		break;					\
+	case 14:					\
+		write_sysreg(val, reg_name##14_el1);	\
+		break;					\
+	case 15:					\
+		write_sysreg(val, reg_name##15_el1);	\
+		break;					\
+	default:					\
+		GUEST_ASSERT(0);			\
+	}						\
+}
+
+/* Define write_dbgbcr()/write_dbgbvr()/write_dbgwcr()/write_dbgwvr() */
+GEN_DEBUG_WRITE_REG(dbgbcr)
+GEN_DEBUG_WRITE_REG(dbgbvr)
+GEN_DEBUG_WRITE_REG(dbgwcr)
+GEN_DEBUG_WRITE_REG(dbgwvr)
+
+static void reset_debug_state(void)
+{
+	uint8_t brps, wrps, i;
+	uint64_t dfr0;
+
+	asm volatile("msr daifset, #8");
+
+	write_sysreg(0, osdlr_el1);
+	write_sysreg(0, oslar_el1);
+	isb();
+
+	write_sysreg(0, mdscr_el1);
+	write_sysreg(0, contextidr_el1);
+
+	/* Reset all bcr/bvr/wcr/wvr registers */
+	dfr0 = read_sysreg(id_aa64dfr0_el1);
+	brps = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_BRPS), dfr0);
+	for (i = 0; i <= brps; i++) {
+		write_dbgbcr(i, 0);
+		write_dbgbvr(i, 0);
+	}
+	wrps = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_WRPS), dfr0);
+	for (i = 0; i <= wrps; i++) {
+		write_dbgwcr(i, 0);
+		write_dbgwvr(i, 0);
+	}
+
+	isb();
+}
+
+static void enable_os_lock(void)
+{
+	write_sysreg(1, oslar_el1);
+	isb();
+
+	GUEST_ASSERT(read_sysreg(oslsr_el1) & 2);
+}
+
+static void enable_monitor_debug_exceptions(void)
+{
+	uint32_t mdscr;
+
+	asm volatile("msr daifclr, #8");
+
+	mdscr = read_sysreg(mdscr_el1) | MDSCR_KDE | MDSCR_MDE;
+	write_sysreg(mdscr, mdscr_el1);
+	isb();
+}
+
+static void install_wp(uint8_t wpn, uint64_t addr)
+{
+	uint32_t wcr;
+
+	wcr = DBGWCR_LEN8 | DBGWCR_RD | DBGWCR_WR | DBGWCR_EL1 | DBGWCR_E;
+	write_dbgwcr(wpn, wcr);
+	write_dbgwvr(wpn, addr);
+
+	isb();
+
+	enable_monitor_debug_exceptions();
+}
+
+static void install_hw_bp(uint8_t bpn, uint64_t addr)
+{
+	uint32_t bcr;
+
+	bcr = DBGBCR_LEN8 | DBGBCR_EXEC | DBGBCR_EL1 | DBGBCR_E;
+	write_dbgbcr(bpn, bcr);
+	write_dbgbvr(bpn, addr);
+	isb();
+
+	enable_monitor_debug_exceptions();
+}
+
+static void install_wp_ctx(uint8_t addr_wp, uint8_t ctx_bp, uint64_t addr,
+			   uint64_t ctx)
+{
+	uint32_t wcr;
+	uint64_t ctx_bcr;
+
+	/* Setup a context-aware breakpoint for Linked Context ID Match */
+	ctx_bcr = DBGBCR_LEN8 | DBGBCR_EXEC | DBGBCR_EL1 | DBGBCR_E |
+		  DBGBCR_BT_CTX_LINK;
+	write_dbgbcr(ctx_bp, ctx_bcr);
+	write_dbgbvr(ctx_bp, ctx);
+
+	/* Setup a linked watchpoint (linked to the context-aware breakpoint) */
+	wcr = DBGWCR_LEN8 | DBGWCR_RD | DBGWCR_WR | DBGWCR_EL1 | DBGWCR_E |
+	      DBGWCR_WT_LINK | ((uint32_t)ctx_bp << DBGWCR_LBN_SHIFT);
+	write_dbgwcr(addr_wp, wcr);
+	write_dbgwvr(addr_wp, addr);
+	isb();
+
+	enable_monitor_debug_exceptions();
+}
+
+void install_hw_bp_ctx(uint8_t addr_bp, uint8_t ctx_bp, uint64_t addr,
+		       uint64_t ctx)
+{
+	uint32_t addr_bcr, ctx_bcr;
+
+	/* Setup a context-aware breakpoint for Linked Context ID Match */
+	ctx_bcr = DBGBCR_LEN8 | DBGBCR_EXEC | DBGBCR_EL1 | DBGBCR_E |
+		  DBGBCR_BT_CTX_LINK;
+	write_dbgbcr(ctx_bp, ctx_bcr);
+	write_dbgbvr(ctx_bp, ctx);
+
+	/*
+	 * Setup a normal breakpoint for Linked Address Match, and link it
+	 * to the context-aware breakpoint.
+	 */
+	addr_bcr = DBGBCR_LEN8 | DBGBCR_EXEC | DBGBCR_EL1 | DBGBCR_E |
+		   DBGBCR_BT_ADDR_LINK_CTX |
+		   ((uint32_t)ctx_bp << DBGBCR_LBN_SHIFT);
+	write_dbgbcr(addr_bp, addr_bcr);
+	write_dbgbvr(addr_bp, addr);
+	isb();
+
+	enable_monitor_debug_exceptions();
+}
+
+static void install_ss(void)
+{
+	uint32_t mdscr;
+
+	asm volatile("msr daifclr, #8");
+
+	mdscr = read_sysreg(mdscr_el1) | MDSCR_KDE | MDSCR_SS;
+	write_sysreg(mdscr, mdscr_el1);
+	isb();
+}
+
+static volatile char write_data;
+
+static void guest_code(uint8_t bpn, uint8_t wpn, uint8_t ctx_bpn)
+{
+	uint64_t ctx = 0xabcdef;	/* a random context number */
+
+	/* Software-breakpoint */
+	reset_debug_state();
+	asm volatile("sw_bp: brk #0");
+	GUEST_ASSERT_EQ(sw_bp_addr, PC(sw_bp));
+
+	/* Hardware-breakpoint */
+	reset_debug_state();
+	install_hw_bp(bpn, PC(hw_bp));
+	asm volatile("hw_bp: nop");
+	GUEST_ASSERT_EQ(hw_bp_addr, PC(hw_bp));
+
+	/* Hardware-breakpoint + svc */
+	reset_debug_state();
+	install_hw_bp(bpn, PC(bp_svc));
+	asm volatile("bp_svc: svc #0");
+	GUEST_ASSERT_EQ(hw_bp_addr, PC(bp_svc));
+	GUEST_ASSERT_EQ(svc_addr, PC(bp_svc) + 4);
+
+	/* Hardware-breakpoint + software-breakpoint */
+	reset_debug_state();
+	install_hw_bp(bpn, PC(bp_brk));
+	asm volatile("bp_brk: brk #0");
+	GUEST_ASSERT_EQ(sw_bp_addr, PC(bp_brk));
+	GUEST_ASSERT_EQ(hw_bp_addr, PC(bp_brk));
+
+	/* Watchpoint */
+	reset_debug_state();
+	install_wp(wpn, PC(write_data));
+	write_data = 'x';
+	GUEST_ASSERT_EQ(write_data, 'x');
+	GUEST_ASSERT_EQ(wp_data_addr, PC(write_data));
+
+	/* Single-step */
+	reset_debug_state();
+	install_ss();
+	ss_idx = 0;
+	asm volatile("ss_start:\n"
+		     "mrs x0, esr_el1\n"
+		     "add x0, x0, #1\n"
+		     "msr daifset, #8\n"
+		     : : : "x0");
+	GUEST_ASSERT_EQ(ss_addr[0], PC(ss_start));
+	GUEST_ASSERT_EQ(ss_addr[1], PC(ss_start) + 4);
+	GUEST_ASSERT_EQ(ss_addr[2], PC(ss_start) + 8);
+
+	/* OS Lock does not block software-breakpoint */
+	reset_debug_state();
+	enable_os_lock();
+	sw_bp_addr = 0;
+	asm volatile("sw_bp2: brk #0");
+	GUEST_ASSERT_EQ(sw_bp_addr, PC(sw_bp2));
+
+	/* OS Lock blocking hardware-breakpoint */
+	reset_debug_state();
+	enable_os_lock();
+	install_hw_bp(bpn, PC(hw_bp2));
+	hw_bp_addr = 0;
+	asm volatile("hw_bp2: nop");
+	GUEST_ASSERT_EQ(hw_bp_addr, 0);
+
+	/* OS Lock blocking watchpoint */
+	reset_debug_state();
+	enable_os_lock();
+	write_data = '\0';
+	wp_data_addr = 0;
+	install_wp(wpn, PC(write_data));
+	write_data = 'x';
+	GUEST_ASSERT_EQ(write_data, 'x');
+	GUEST_ASSERT_EQ(wp_data_addr, 0);
+
+	/* OS Lock blocking single-step */
+	reset_debug_state();
+	enable_os_lock();
+	ss_addr[0] = 0;
+	install_ss();
+	ss_idx = 0;
+	asm volatile("mrs x0, esr_el1\n\t"
+		     "add x0, x0, #1\n\t"
+		     "msr daifset, #8\n\t"
+		     : : : "x0");
+	GUEST_ASSERT_EQ(ss_addr[0], 0);
+
+	/* Linked hardware-breakpoint */
+	hw_bp_addr = 0;
+	reset_debug_state();
+	install_hw_bp_ctx(bpn, ctx_bpn, PC(hw_bp_ctx), ctx);
+	/* Set context id */
+	write_sysreg(ctx, contextidr_el1);
+	isb();
+	asm volatile("hw_bp_ctx: nop");
+	write_sysreg(0, contextidr_el1);
+	GUEST_ASSERT_EQ(hw_bp_addr, PC(hw_bp_ctx));
+
+	/* Linked watchpoint */
+	reset_debug_state();
+	install_wp_ctx(wpn, ctx_bpn, PC(write_data), ctx);
+	/* Set context id */
+	write_sysreg(ctx, contextidr_el1);
+	isb();
+	write_data = 'x';
+	GUEST_ASSERT_EQ(write_data, 'x');
+	GUEST_ASSERT_EQ(wp_data_addr, PC(write_data));
+
+	GUEST_DONE();
+}
+
+static void guest_sw_bp_handler(struct ex_regs *regs)
+{
+	sw_bp_addr = regs->pc;
+	regs->pc += 4;
+}
+
+static void guest_hw_bp_handler(struct ex_regs *regs)
+{
+	hw_bp_addr = regs->pc;
+	regs->pstate |= SPSR_D;
+}
+
+static void guest_wp_handler(struct ex_regs *regs)
+{
+	wp_data_addr = read_sysreg(far_el1);
+	wp_addr = regs->pc;
+	regs->pstate |= SPSR_D;
+}
+
+static void guest_ss_handler(struct ex_regs *regs)
+{
+	__GUEST_ASSERT(ss_idx < 4, "Expected index < 4, got '%u'", ss_idx);
+	ss_addr[ss_idx++] = regs->pc;
+	regs->pstate |= SPSR_SS;
+}
+
+static void guest_svc_handler(struct ex_regs *regs)
+{
+	svc_addr = regs->pc;
+}
+
+static void guest_code_ss(int test_cnt)
+{
+	uint64_t i;
+	uint64_t bvr, wvr, w_bvr, w_wvr;
+
+	for (i = 0; i < test_cnt; i++) {
+		/* Bits [1:0] of dbg{b,w}vr are RES0 */
+		w_bvr = i << 2;
+		w_wvr = i << 2;
+
+		/*
+		 * Enable Single Step execution.  Note!  This _must_ be a bare
+		 * ucall as the ucall() path uses atomic operations to manage
+		 * the ucall structures, and the built-in "atomics" are usually
+		 * implemented via exclusive access instructions.  The exlusive
+		 * monitor is cleared on ERET, and so taking debug exceptions
+		 * during a LDREX=>STREX sequence will prevent forward progress
+		 * and hang the guest/test.
+		 */
+		GUEST_UCALL_NONE();
+
+		/*
+		 * The userspace will verify that the pc is as expected during
+		 * single step execution between iter_ss_begin and iter_ss_end.
+		 */
+		asm volatile("iter_ss_begin:nop\n");
+
+		write_sysreg(w_bvr, dbgbvr0_el1);
+		write_sysreg(w_wvr, dbgwvr0_el1);
+		bvr = read_sysreg(dbgbvr0_el1);
+		wvr = read_sysreg(dbgwvr0_el1);
+
+		/* Userspace disables Single Step when the end is nigh. */
+		asm volatile("iter_ss_end:\n");
+
+		GUEST_ASSERT_EQ(bvr, w_bvr);
+		GUEST_ASSERT_EQ(wvr, w_wvr);
+	}
+	GUEST_DONE();
+}
+
+static int debug_version(uint64_t id_aa64dfr0)
+{
+	return FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_DEBUGVER), id_aa64dfr0);
+}
+
+static void test_guest_debug_exceptions(uint8_t bpn, uint8_t wpn, uint8_t ctx_bpn)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	struct ucall uc;
+
+	vm = vm_create_with_one_vcpu(&vcpu, guest_code);
+
+	vm_init_descriptor_tables(vm);
+	vcpu_init_descriptor_tables(vcpu);
+
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_BRK_INS, guest_sw_bp_handler);
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_HW_BP_CURRENT, guest_hw_bp_handler);
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_WP_CURRENT, guest_wp_handler);
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_SSTEP_CURRENT, guest_ss_handler);
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_SVC64, guest_svc_handler);
+
+	/* Specify bpn/wpn/ctx_bpn to be tested */
+	vcpu_args_set(vcpu, 3, bpn, wpn, ctx_bpn);
+	pr_debug("Use bpn#%d, wpn#%d and ctx_bpn#%d\n", bpn, wpn, ctx_bpn);
+
+	vcpu_run(vcpu);
+	switch (get_ucall(vcpu, &uc)) {
+	case UCALL_ABORT:
+		REPORT_GUEST_ASSERT(uc);
+		break;
+	case UCALL_DONE:
+		goto done;
+	default:
+		TEST_FAIL("Unknown ucall %lu", uc.cmd);
+	}
+
+done:
+	kvm_vm_free(vm);
+}
+
+void test_single_step_from_userspace(int test_cnt)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	struct ucall uc;
+	struct kvm_run *run;
+	uint64_t pc, cmd;
+	uint64_t test_pc = 0;
+	bool ss_enable = false;
+	struct kvm_guest_debug debug = {};
+
+	vm = vm_create_with_one_vcpu(&vcpu, guest_code_ss);
+	run = vcpu->run;
+	vcpu_args_set(vcpu, 1, test_cnt);
+
+	while (1) {
+		vcpu_run(vcpu);
+		if (run->exit_reason != KVM_EXIT_DEBUG) {
+			cmd = get_ucall(vcpu, &uc);
+			if (cmd == UCALL_ABORT) {
+				REPORT_GUEST_ASSERT(uc);
+				/* NOT REACHED */
+			} else if (cmd == UCALL_DONE) {
+				break;
+			}
+
+			TEST_ASSERT(cmd == UCALL_NONE,
+				    "Unexpected ucall cmd 0x%lx", cmd);
+
+			debug.control = KVM_GUESTDBG_ENABLE |
+					KVM_GUESTDBG_SINGLESTEP;
+			ss_enable = true;
+			vcpu_guest_debug_set(vcpu, &debug);
+			continue;
+		}
+
+		TEST_ASSERT(ss_enable, "Unexpected KVM_EXIT_DEBUG");
+
+		/* Check if the current pc is expected. */
+		vcpu_get_reg(vcpu, ARM64_CORE_REG(regs.pc), &pc);
+		TEST_ASSERT(!test_pc || pc == test_pc,
+			    "Unexpected pc 0x%lx (expected 0x%lx)",
+			    pc, test_pc);
+
+		if ((pc + 4) == (uint64_t)&iter_ss_end) {
+			test_pc = 0;
+			debug.control = KVM_GUESTDBG_ENABLE;
+			ss_enable = false;
+			vcpu_guest_debug_set(vcpu, &debug);
+			continue;
+		}
+
+		/*
+		 * If the current pc is between iter_ss_bgin and
+		 * iter_ss_end, the pc for the next KVM_EXIT_DEBUG should
+		 * be the current pc + 4.
+		 */
+		if ((pc >= (uint64_t)&iter_ss_begin) &&
+		    (pc < (uint64_t)&iter_ss_end))
+			test_pc = pc + 4;
+		else
+			test_pc = 0;
+	}
+
+	kvm_vm_free(vm);
+}
+
+/*
+ * Run debug testing using the various breakpoint#, watchpoint# and
+ * context-aware breakpoint# with the given ID_AA64DFR0_EL1 configuration.
+ */
+void test_guest_debug_exceptions_all(uint64_t aa64dfr0)
+{
+	uint8_t brp_num, wrp_num, ctx_brp_num, normal_brp_num, ctx_brp_base;
+	int b, w, c;
+
+	/* Number of breakpoints */
+	brp_num = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_BRPS), aa64dfr0) + 1;
+	__TEST_REQUIRE(brp_num >= 2, "At least two breakpoints are required");
+
+	/* Number of watchpoints */
+	wrp_num = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_WRPS), aa64dfr0) + 1;
+
+	/* Number of context aware breakpoints */
+	ctx_brp_num = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64DFR0_CTX_CMPS), aa64dfr0) + 1;
+
+	pr_debug("%s brp_num:%d, wrp_num:%d, ctx_brp_num:%d\n", __func__,
+		 brp_num, wrp_num, ctx_brp_num);
+
+	/* Number of normal (non-context aware) breakpoints */
+	normal_brp_num = brp_num - ctx_brp_num;
+
+	/* Lowest context aware breakpoint number */
+	ctx_brp_base = normal_brp_num;
+
+	/* Run tests with all supported breakpoints/watchpoints */
+	for (c = ctx_brp_base; c < ctx_brp_base + ctx_brp_num; c++) {
+		for (b = 0; b < normal_brp_num; b++) {
+			for (w = 0; w < wrp_num; w++)
+				test_guest_debug_exceptions(b, w, c);
+		}
+	}
+}
+
+static void help(char *name)
+{
+	puts("");
+	printf("Usage: %s [-h] [-i iterations of the single step test]\n", name);
+	puts("");
+	exit(0);
+}
+
+int main(int argc, char *argv[])
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	int opt;
+	int ss_iteration = 10000;
+	uint64_t aa64dfr0;
+
+	vm = vm_create_with_one_vcpu(&vcpu, guest_code);
+	vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(SYS_ID_AA64DFR0_EL1), &aa64dfr0);
+	__TEST_REQUIRE(debug_version(aa64dfr0) >= 6,
+		       "Armv8 debug architecture not supported.");
+	kvm_vm_free(vm);
+
+	while ((opt = getopt(argc, argv, "i:")) != -1) {
+		switch (opt) {
+		case 'i':
+			ss_iteration = atoi_positive("Number of iterations", optarg);
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			break;
+		}
+	}
+
+	test_guest_debug_exceptions_all(aa64dfr0);
+	test_single_step_from_userspace(ss_iteration);
+
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/get-reg-list.c b/tools/testing/selftests/kvm/aarch64/get-reg-list.c
new file mode 100644
index 0000000000..709d7d7217
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/get-reg-list.c
@@ -0,0 +1,757 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Check for KVM_GET_REG_LIST regressions.
+ *
+ * Copyright (C) 2020, Red Hat, Inc.
+ *
+ * While the blessed list should be created from the oldest possible
+ * kernel, we can't go older than v5.2, though, because that's the first
+ * release which includes df205b5c6328 ("KVM: arm64: Filter out invalid
+ * core register IDs in KVM_GET_REG_LIST"). Without that commit the core
+ * registers won't match expectations.
+ */
+#include <stdio.h>
+#include "kvm_util.h"
+#include "test_util.h"
+#include "processor.h"
+
+struct feature_id_reg {
+	__u64 reg;
+	__u64 id_reg;
+	__u64 feat_shift;
+	__u64 feat_min;
+};
+
+static struct feature_id_reg feat_id_regs[] = {
+	{
+		ARM64_SYS_REG(3, 0, 2, 0, 3),	/* TCR2_EL1 */
+		ARM64_SYS_REG(3, 0, 0, 7, 3),	/* ID_AA64MMFR3_EL1 */
+		0,
+		1
+	},
+	{
+		ARM64_SYS_REG(3, 0, 10, 2, 2),	/* PIRE0_EL1 */
+		ARM64_SYS_REG(3, 0, 0, 7, 3),	/* ID_AA64MMFR3_EL1 */
+		4,
+		1
+	},
+	{
+		ARM64_SYS_REG(3, 0, 10, 2, 3),	/* PIR_EL1 */
+		ARM64_SYS_REG(3, 0, 0, 7, 3),	/* ID_AA64MMFR3_EL1 */
+		4,
+		1
+	}
+};
+
+bool filter_reg(__u64 reg)
+{
+	/*
+	 * DEMUX register presence depends on the host's CLIDR_EL1.
+	 * This means there's no set of them that we can bless.
+	 */
+	if ((reg & KVM_REG_ARM_COPROC_MASK) == KVM_REG_ARM_DEMUX)
+		return true;
+
+	return false;
+}
+
+static bool check_supported_feat_reg(struct kvm_vcpu *vcpu, __u64 reg)
+{
+	int i, ret;
+	__u64 data, feat_val;
+
+	for (i = 0; i < ARRAY_SIZE(feat_id_regs); i++) {
+		if (feat_id_regs[i].reg == reg) {
+			ret = __vcpu_get_reg(vcpu, feat_id_regs[i].id_reg, &data);
+			if (ret < 0)
+				return false;
+
+			feat_val = ((data >> feat_id_regs[i].feat_shift) & 0xf);
+			return feat_val >= feat_id_regs[i].feat_min;
+		}
+	}
+
+	return true;
+}
+
+bool check_supported_reg(struct kvm_vcpu *vcpu, __u64 reg)
+{
+	return check_supported_feat_reg(vcpu, reg);
+}
+
+bool check_reject_set(int err)
+{
+	return err == EPERM;
+}
+
+void finalize_vcpu(struct kvm_vcpu *vcpu, struct vcpu_reg_list *c)
+{
+	struct vcpu_reg_sublist *s;
+	int feature;
+
+	for_each_sublist(c, s) {
+		if (s->finalize) {
+			feature = s->feature;
+			vcpu_ioctl(vcpu, KVM_ARM_VCPU_FINALIZE, &feature);
+		}
+	}
+}
+
+#define REG_MASK (KVM_REG_ARCH_MASK | KVM_REG_SIZE_MASK | KVM_REG_ARM_COPROC_MASK)
+
+#define CORE_REGS_XX_NR_WORDS	2
+#define CORE_SPSR_XX_NR_WORDS	2
+#define CORE_FPREGS_XX_NR_WORDS	4
+
+static const char *core_id_to_str(const char *prefix, __u64 id)
+{
+	__u64 core_off = id & ~REG_MASK, idx;
+
+	/*
+	 * core_off is the offset into struct kvm_regs
+	 */
+	switch (core_off) {
+	case KVM_REG_ARM_CORE_REG(regs.regs[0]) ...
+	     KVM_REG_ARM_CORE_REG(regs.regs[30]):
+		idx = (core_off - KVM_REG_ARM_CORE_REG(regs.regs[0])) / CORE_REGS_XX_NR_WORDS;
+		TEST_ASSERT(idx < 31, "%s: Unexpected regs.regs index: %lld", prefix, idx);
+		return strdup_printf("KVM_REG_ARM_CORE_REG(regs.regs[%lld])", idx);
+	case KVM_REG_ARM_CORE_REG(regs.sp):
+		return "KVM_REG_ARM_CORE_REG(regs.sp)";
+	case KVM_REG_ARM_CORE_REG(regs.pc):
+		return "KVM_REG_ARM_CORE_REG(regs.pc)";
+	case KVM_REG_ARM_CORE_REG(regs.pstate):
+		return "KVM_REG_ARM_CORE_REG(regs.pstate)";
+	case KVM_REG_ARM_CORE_REG(sp_el1):
+		return "KVM_REG_ARM_CORE_REG(sp_el1)";
+	case KVM_REG_ARM_CORE_REG(elr_el1):
+		return "KVM_REG_ARM_CORE_REG(elr_el1)";
+	case KVM_REG_ARM_CORE_REG(spsr[0]) ...
+	     KVM_REG_ARM_CORE_REG(spsr[KVM_NR_SPSR - 1]):
+		idx = (core_off - KVM_REG_ARM_CORE_REG(spsr[0])) / CORE_SPSR_XX_NR_WORDS;
+		TEST_ASSERT(idx < KVM_NR_SPSR, "%s: Unexpected spsr index: %lld", prefix, idx);
+		return strdup_printf("KVM_REG_ARM_CORE_REG(spsr[%lld])", idx);
+	case KVM_REG_ARM_CORE_REG(fp_regs.vregs[0]) ...
+	     KVM_REG_ARM_CORE_REG(fp_regs.vregs[31]):
+		idx = (core_off - KVM_REG_ARM_CORE_REG(fp_regs.vregs[0])) / CORE_FPREGS_XX_NR_WORDS;
+		TEST_ASSERT(idx < 32, "%s: Unexpected fp_regs.vregs index: %lld", prefix, idx);
+		return strdup_printf("KVM_REG_ARM_CORE_REG(fp_regs.vregs[%lld])", idx);
+	case KVM_REG_ARM_CORE_REG(fp_regs.fpsr):
+		return "KVM_REG_ARM_CORE_REG(fp_regs.fpsr)";
+	case KVM_REG_ARM_CORE_REG(fp_regs.fpcr):
+		return "KVM_REG_ARM_CORE_REG(fp_regs.fpcr)";
+	}
+
+	TEST_FAIL("%s: Unknown core reg id: 0x%llx", prefix, id);
+	return NULL;
+}
+
+static const char *sve_id_to_str(const char *prefix, __u64 id)
+{
+	__u64 sve_off, n, i;
+
+	if (id == KVM_REG_ARM64_SVE_VLS)
+		return "KVM_REG_ARM64_SVE_VLS";
+
+	sve_off = id & ~(REG_MASK | ((1ULL << 5) - 1));
+	i = id & (KVM_ARM64_SVE_MAX_SLICES - 1);
+
+	TEST_ASSERT(i == 0, "%s: Currently we don't expect slice > 0, reg id 0x%llx", prefix, id);
+
+	switch (sve_off) {
+	case KVM_REG_ARM64_SVE_ZREG_BASE ...
+	     KVM_REG_ARM64_SVE_ZREG_BASE + (1ULL << 5) * KVM_ARM64_SVE_NUM_ZREGS - 1:
+		n = (id >> 5) & (KVM_ARM64_SVE_NUM_ZREGS - 1);
+		TEST_ASSERT(id == KVM_REG_ARM64_SVE_ZREG(n, 0),
+			    "%s: Unexpected bits set in SVE ZREG id: 0x%llx", prefix, id);
+		return strdup_printf("KVM_REG_ARM64_SVE_ZREG(%lld, 0)", n);
+	case KVM_REG_ARM64_SVE_PREG_BASE ...
+	     KVM_REG_ARM64_SVE_PREG_BASE + (1ULL << 5) * KVM_ARM64_SVE_NUM_PREGS - 1:
+		n = (id >> 5) & (KVM_ARM64_SVE_NUM_PREGS - 1);
+		TEST_ASSERT(id == KVM_REG_ARM64_SVE_PREG(n, 0),
+			    "%s: Unexpected bits set in SVE PREG id: 0x%llx", prefix, id);
+		return strdup_printf("KVM_REG_ARM64_SVE_PREG(%lld, 0)", n);
+	case KVM_REG_ARM64_SVE_FFR_BASE:
+		TEST_ASSERT(id == KVM_REG_ARM64_SVE_FFR(0),
+			    "%s: Unexpected bits set in SVE FFR id: 0x%llx", prefix, id);
+		return "KVM_REG_ARM64_SVE_FFR(0)";
+	}
+
+	return NULL;
+}
+
+void print_reg(const char *prefix, __u64 id)
+{
+	unsigned op0, op1, crn, crm, op2;
+	const char *reg_size = NULL;
+
+	TEST_ASSERT((id & KVM_REG_ARCH_MASK) == KVM_REG_ARM64,
+		    "%s: KVM_REG_ARM64 missing in reg id: 0x%llx", prefix, id);
+
+	switch (id & KVM_REG_SIZE_MASK) {
+	case KVM_REG_SIZE_U8:
+		reg_size = "KVM_REG_SIZE_U8";
+		break;
+	case KVM_REG_SIZE_U16:
+		reg_size = "KVM_REG_SIZE_U16";
+		break;
+	case KVM_REG_SIZE_U32:
+		reg_size = "KVM_REG_SIZE_U32";
+		break;
+	case KVM_REG_SIZE_U64:
+		reg_size = "KVM_REG_SIZE_U64";
+		break;
+	case KVM_REG_SIZE_U128:
+		reg_size = "KVM_REG_SIZE_U128";
+		break;
+	case KVM_REG_SIZE_U256:
+		reg_size = "KVM_REG_SIZE_U256";
+		break;
+	case KVM_REG_SIZE_U512:
+		reg_size = "KVM_REG_SIZE_U512";
+		break;
+	case KVM_REG_SIZE_U1024:
+		reg_size = "KVM_REG_SIZE_U1024";
+		break;
+	case KVM_REG_SIZE_U2048:
+		reg_size = "KVM_REG_SIZE_U2048";
+		break;
+	default:
+		TEST_FAIL("%s: Unexpected reg size: 0x%llx in reg id: 0x%llx",
+			  prefix, (id & KVM_REG_SIZE_MASK) >> KVM_REG_SIZE_SHIFT, id);
+	}
+
+	switch (id & KVM_REG_ARM_COPROC_MASK) {
+	case KVM_REG_ARM_CORE:
+		printf("\tKVM_REG_ARM64 | %s | KVM_REG_ARM_CORE | %s,\n", reg_size, core_id_to_str(prefix, id));
+		break;
+	case KVM_REG_ARM_DEMUX:
+		TEST_ASSERT(!(id & ~(REG_MASK | KVM_REG_ARM_DEMUX_ID_MASK | KVM_REG_ARM_DEMUX_VAL_MASK)),
+			    "%s: Unexpected bits set in DEMUX reg id: 0x%llx", prefix, id);
+		printf("\tKVM_REG_ARM64 | %s | KVM_REG_ARM_DEMUX | KVM_REG_ARM_DEMUX_ID_CCSIDR | %lld,\n",
+		       reg_size, id & KVM_REG_ARM_DEMUX_VAL_MASK);
+		break;
+	case KVM_REG_ARM64_SYSREG:
+		op0 = (id & KVM_REG_ARM64_SYSREG_OP0_MASK) >> KVM_REG_ARM64_SYSREG_OP0_SHIFT;
+		op1 = (id & KVM_REG_ARM64_SYSREG_OP1_MASK) >> KVM_REG_ARM64_SYSREG_OP1_SHIFT;
+		crn = (id & KVM_REG_ARM64_SYSREG_CRN_MASK) >> KVM_REG_ARM64_SYSREG_CRN_SHIFT;
+		crm = (id & KVM_REG_ARM64_SYSREG_CRM_MASK) >> KVM_REG_ARM64_SYSREG_CRM_SHIFT;
+		op2 = (id & KVM_REG_ARM64_SYSREG_OP2_MASK) >> KVM_REG_ARM64_SYSREG_OP2_SHIFT;
+		TEST_ASSERT(id == ARM64_SYS_REG(op0, op1, crn, crm, op2),
+			    "%s: Unexpected bits set in SYSREG reg id: 0x%llx", prefix, id);
+		printf("\tARM64_SYS_REG(%d, %d, %d, %d, %d),\n", op0, op1, crn, crm, op2);
+		break;
+	case KVM_REG_ARM_FW:
+		TEST_ASSERT(id == KVM_REG_ARM_FW_REG(id & 0xffff),
+			    "%s: Unexpected bits set in FW reg id: 0x%llx", prefix, id);
+		printf("\tKVM_REG_ARM_FW_REG(%lld),\n", id & 0xffff);
+		break;
+	case KVM_REG_ARM_FW_FEAT_BMAP:
+		TEST_ASSERT(id == KVM_REG_ARM_FW_FEAT_BMAP_REG(id & 0xffff),
+			    "%s: Unexpected bits set in the bitmap feature FW reg id: 0x%llx", prefix, id);
+		printf("\tKVM_REG_ARM_FW_FEAT_BMAP_REG(%lld),\n", id & 0xffff);
+		break;
+	case KVM_REG_ARM64_SVE:
+		printf("\t%s,\n", sve_id_to_str(prefix, id));
+		break;
+	default:
+		TEST_FAIL("%s: Unexpected coproc type: 0x%llx in reg id: 0x%llx",
+			  prefix, (id & KVM_REG_ARM_COPROC_MASK) >> KVM_REG_ARM_COPROC_SHIFT, id);
+	}
+}
+
+/*
+ * The original blessed list was primed with the output of kernel version
+ * v4.15 with --core-reg-fixup and then later updated with new registers.
+ * (The --core-reg-fixup option and it's fixup function have been removed
+ * from the test, as it's unlikely to use this type of test on a kernel
+ * older than v5.2.)
+ *
+ * The blessed list is up to date with kernel version v6.4 (or so we hope)
+ */
+static __u64 base_regs[] = {
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[0]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[1]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[2]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[3]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[4]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[5]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[6]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[7]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[8]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[9]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[10]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[11]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[12]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[13]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[14]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[15]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[16]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[17]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[18]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[19]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[20]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[21]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[22]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[23]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[24]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[25]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[26]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[27]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[28]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[29]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.regs[30]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.sp),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.pc),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(regs.pstate),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(sp_el1),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(elr_el1),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[0]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[1]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[2]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[3]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[4]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U32 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.fpsr),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U32 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.fpcr),
+	KVM_REG_ARM_FW_REG(0),		/* KVM_REG_ARM_PSCI_VERSION */
+	KVM_REG_ARM_FW_REG(1),		/* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_1 */
+	KVM_REG_ARM_FW_REG(2),		/* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_2 */
+	KVM_REG_ARM_FW_REG(3),		/* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_3 */
+	KVM_REG_ARM_FW_FEAT_BMAP_REG(0),	/* KVM_REG_ARM_STD_BMAP */
+	KVM_REG_ARM_FW_FEAT_BMAP_REG(1),	/* KVM_REG_ARM_STD_HYP_BMAP */
+	KVM_REG_ARM_FW_FEAT_BMAP_REG(2),	/* KVM_REG_ARM_VENDOR_HYP_BMAP */
+	ARM64_SYS_REG(3, 3, 14, 3, 1),	/* CNTV_CTL_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 3, 2),	/* CNTV_CVAL_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 0, 2),
+	ARM64_SYS_REG(3, 0, 0, 0, 0),	/* MIDR_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 0, 6),	/* REVIDR_EL1 */
+	ARM64_SYS_REG(3, 1, 0, 0, 1),	/* CLIDR_EL1 */
+	ARM64_SYS_REG(3, 1, 0, 0, 7),	/* AIDR_EL1 */
+	ARM64_SYS_REG(3, 3, 0, 0, 1),	/* CTR_EL0 */
+	ARM64_SYS_REG(2, 0, 0, 0, 4),
+	ARM64_SYS_REG(2, 0, 0, 0, 5),
+	ARM64_SYS_REG(2, 0, 0, 0, 6),
+	ARM64_SYS_REG(2, 0, 0, 0, 7),
+	ARM64_SYS_REG(2, 0, 0, 1, 4),
+	ARM64_SYS_REG(2, 0, 0, 1, 5),
+	ARM64_SYS_REG(2, 0, 0, 1, 6),
+	ARM64_SYS_REG(2, 0, 0, 1, 7),
+	ARM64_SYS_REG(2, 0, 0, 2, 0),	/* MDCCINT_EL1 */
+	ARM64_SYS_REG(2, 0, 0, 2, 2),	/* MDSCR_EL1 */
+	ARM64_SYS_REG(2, 0, 0, 2, 4),
+	ARM64_SYS_REG(2, 0, 0, 2, 5),
+	ARM64_SYS_REG(2, 0, 0, 2, 6),
+	ARM64_SYS_REG(2, 0, 0, 2, 7),
+	ARM64_SYS_REG(2, 0, 0, 3, 4),
+	ARM64_SYS_REG(2, 0, 0, 3, 5),
+	ARM64_SYS_REG(2, 0, 0, 3, 6),
+	ARM64_SYS_REG(2, 0, 0, 3, 7),
+	ARM64_SYS_REG(2, 0, 0, 4, 4),
+	ARM64_SYS_REG(2, 0, 0, 4, 5),
+	ARM64_SYS_REG(2, 0, 0, 4, 6),
+	ARM64_SYS_REG(2, 0, 0, 4, 7),
+	ARM64_SYS_REG(2, 0, 0, 5, 4),
+	ARM64_SYS_REG(2, 0, 0, 5, 5),
+	ARM64_SYS_REG(2, 0, 0, 5, 6),
+	ARM64_SYS_REG(2, 0, 0, 5, 7),
+	ARM64_SYS_REG(2, 0, 0, 6, 4),
+	ARM64_SYS_REG(2, 0, 0, 6, 5),
+	ARM64_SYS_REG(2, 0, 0, 6, 6),
+	ARM64_SYS_REG(2, 0, 0, 6, 7),
+	ARM64_SYS_REG(2, 0, 0, 7, 4),
+	ARM64_SYS_REG(2, 0, 0, 7, 5),
+	ARM64_SYS_REG(2, 0, 0, 7, 6),
+	ARM64_SYS_REG(2, 0, 0, 7, 7),
+	ARM64_SYS_REG(2, 0, 0, 8, 4),
+	ARM64_SYS_REG(2, 0, 0, 8, 5),
+	ARM64_SYS_REG(2, 0, 0, 8, 6),
+	ARM64_SYS_REG(2, 0, 0, 8, 7),
+	ARM64_SYS_REG(2, 0, 0, 9, 4),
+	ARM64_SYS_REG(2, 0, 0, 9, 5),
+	ARM64_SYS_REG(2, 0, 0, 9, 6),
+	ARM64_SYS_REG(2, 0, 0, 9, 7),
+	ARM64_SYS_REG(2, 0, 0, 10, 4),
+	ARM64_SYS_REG(2, 0, 0, 10, 5),
+	ARM64_SYS_REG(2, 0, 0, 10, 6),
+	ARM64_SYS_REG(2, 0, 0, 10, 7),
+	ARM64_SYS_REG(2, 0, 0, 11, 4),
+	ARM64_SYS_REG(2, 0, 0, 11, 5),
+	ARM64_SYS_REG(2, 0, 0, 11, 6),
+	ARM64_SYS_REG(2, 0, 0, 11, 7),
+	ARM64_SYS_REG(2, 0, 0, 12, 4),
+	ARM64_SYS_REG(2, 0, 0, 12, 5),
+	ARM64_SYS_REG(2, 0, 0, 12, 6),
+	ARM64_SYS_REG(2, 0, 0, 12, 7),
+	ARM64_SYS_REG(2, 0, 0, 13, 4),
+	ARM64_SYS_REG(2, 0, 0, 13, 5),
+	ARM64_SYS_REG(2, 0, 0, 13, 6),
+	ARM64_SYS_REG(2, 0, 0, 13, 7),
+	ARM64_SYS_REG(2, 0, 0, 14, 4),
+	ARM64_SYS_REG(2, 0, 0, 14, 5),
+	ARM64_SYS_REG(2, 0, 0, 14, 6),
+	ARM64_SYS_REG(2, 0, 0, 14, 7),
+	ARM64_SYS_REG(2, 0, 0, 15, 4),
+	ARM64_SYS_REG(2, 0, 0, 15, 5),
+	ARM64_SYS_REG(2, 0, 0, 15, 6),
+	ARM64_SYS_REG(2, 0, 0, 15, 7),
+	ARM64_SYS_REG(2, 0, 1, 1, 4),	/* OSLSR_EL1 */
+	ARM64_SYS_REG(2, 4, 0, 7, 0),	/* DBGVCR32_EL2 */
+	ARM64_SYS_REG(3, 0, 0, 0, 5),	/* MPIDR_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 0),	/* ID_PFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 1),	/* ID_PFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 2),	/* ID_DFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 3),	/* ID_AFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 4),	/* ID_MMFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 5),	/* ID_MMFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 6),	/* ID_MMFR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 1, 7),	/* ID_MMFR3_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 0),	/* ID_ISAR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 1),	/* ID_ISAR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 2),	/* ID_ISAR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 3),	/* ID_ISAR3_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 4),	/* ID_ISAR4_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 5),	/* ID_ISAR5_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 6),	/* ID_MMFR4_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 2, 7),	/* ID_ISAR6_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 0),	/* MVFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 1),	/* MVFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 2),	/* MVFR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 3),
+	ARM64_SYS_REG(3, 0, 0, 3, 4),	/* ID_PFR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 5),	/* ID_DFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 6),	/* ID_MMFR5_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 3, 7),
+	ARM64_SYS_REG(3, 0, 0, 4, 0),	/* ID_AA64PFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 4, 1),	/* ID_AA64PFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 4, 2),	/* ID_AA64PFR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 4, 3),
+	ARM64_SYS_REG(3, 0, 0, 4, 4),	/* ID_AA64ZFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 4, 5),	/* ID_AA64SMFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 4, 6),
+	ARM64_SYS_REG(3, 0, 0, 4, 7),
+	ARM64_SYS_REG(3, 0, 0, 5, 0),	/* ID_AA64DFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 5, 1),	/* ID_AA64DFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 5, 2),
+	ARM64_SYS_REG(3, 0, 0, 5, 3),
+	ARM64_SYS_REG(3, 0, 0, 5, 4),	/* ID_AA64AFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 5, 5),	/* ID_AA64AFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 5, 6),
+	ARM64_SYS_REG(3, 0, 0, 5, 7),
+	ARM64_SYS_REG(3, 0, 0, 6, 0),	/* ID_AA64ISAR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 6, 1),	/* ID_AA64ISAR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 6, 2),	/* ID_AA64ISAR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 6, 3),
+	ARM64_SYS_REG(3, 0, 0, 6, 4),
+	ARM64_SYS_REG(3, 0, 0, 6, 5),
+	ARM64_SYS_REG(3, 0, 0, 6, 6),
+	ARM64_SYS_REG(3, 0, 0, 6, 7),
+	ARM64_SYS_REG(3, 0, 0, 7, 0),	/* ID_AA64MMFR0_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 7, 1),	/* ID_AA64MMFR1_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 7, 2),	/* ID_AA64MMFR2_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 7, 3),	/* ID_AA64MMFR3_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 7, 4),	/* ID_AA64MMFR4_EL1 */
+	ARM64_SYS_REG(3, 0, 0, 7, 5),
+	ARM64_SYS_REG(3, 0, 0, 7, 6),
+	ARM64_SYS_REG(3, 0, 0, 7, 7),
+	ARM64_SYS_REG(3, 0, 1, 0, 0),	/* SCTLR_EL1 */
+	ARM64_SYS_REG(3, 0, 1, 0, 1),	/* ACTLR_EL1 */
+	ARM64_SYS_REG(3, 0, 1, 0, 2),	/* CPACR_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 0, 0),	/* TTBR0_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 0, 1),	/* TTBR1_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 0, 2),	/* TCR_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 0, 3),	/* TCR2_EL1 */
+	ARM64_SYS_REG(3, 0, 5, 1, 0),	/* AFSR0_EL1 */
+	ARM64_SYS_REG(3, 0, 5, 1, 1),	/* AFSR1_EL1 */
+	ARM64_SYS_REG(3, 0, 5, 2, 0),	/* ESR_EL1 */
+	ARM64_SYS_REG(3, 0, 6, 0, 0),	/* FAR_EL1 */
+	ARM64_SYS_REG(3, 0, 7, 4, 0),	/* PAR_EL1 */
+	ARM64_SYS_REG(3, 0, 10, 2, 0),	/* MAIR_EL1 */
+	ARM64_SYS_REG(3, 0, 10, 2, 2),	/* PIRE0_EL1 */
+	ARM64_SYS_REG(3, 0, 10, 2, 3),	/* PIR_EL1 */
+	ARM64_SYS_REG(3, 0, 10, 3, 0),	/* AMAIR_EL1 */
+	ARM64_SYS_REG(3, 0, 12, 0, 0),	/* VBAR_EL1 */
+	ARM64_SYS_REG(3, 0, 12, 1, 1),	/* DISR_EL1 */
+	ARM64_SYS_REG(3, 0, 13, 0, 1),	/* CONTEXTIDR_EL1 */
+	ARM64_SYS_REG(3, 0, 13, 0, 4),	/* TPIDR_EL1 */
+	ARM64_SYS_REG(3, 0, 14, 1, 0),	/* CNTKCTL_EL1 */
+	ARM64_SYS_REG(3, 2, 0, 0, 0),	/* CSSELR_EL1 */
+	ARM64_SYS_REG(3, 3, 13, 0, 2),	/* TPIDR_EL0 */
+	ARM64_SYS_REG(3, 3, 13, 0, 3),	/* TPIDRRO_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 0, 1),	/* CNTPCT_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 2, 1),	/* CNTP_CTL_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 2, 2),	/* CNTP_CVAL_EL0 */
+	ARM64_SYS_REG(3, 4, 3, 0, 0),	/* DACR32_EL2 */
+	ARM64_SYS_REG(3, 4, 5, 0, 1),	/* IFSR32_EL2 */
+	ARM64_SYS_REG(3, 4, 5, 3, 0),	/* FPEXC32_EL2 */
+};
+
+static __u64 pmu_regs[] = {
+	ARM64_SYS_REG(3, 0, 9, 14, 1),	/* PMINTENSET_EL1 */
+	ARM64_SYS_REG(3, 0, 9, 14, 2),	/* PMINTENCLR_EL1 */
+	ARM64_SYS_REG(3, 3, 9, 12, 0),	/* PMCR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 12, 1),	/* PMCNTENSET_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 12, 2),	/* PMCNTENCLR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 12, 3),	/* PMOVSCLR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 12, 4),	/* PMSWINC_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 12, 5),	/* PMSELR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 13, 0),	/* PMCCNTR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 14, 0),	/* PMUSERENR_EL0 */
+	ARM64_SYS_REG(3, 3, 9, 14, 3),	/* PMOVSSET_EL0 */
+	ARM64_SYS_REG(3, 3, 14, 8, 0),
+	ARM64_SYS_REG(3, 3, 14, 8, 1),
+	ARM64_SYS_REG(3, 3, 14, 8, 2),
+	ARM64_SYS_REG(3, 3, 14, 8, 3),
+	ARM64_SYS_REG(3, 3, 14, 8, 4),
+	ARM64_SYS_REG(3, 3, 14, 8, 5),
+	ARM64_SYS_REG(3, 3, 14, 8, 6),
+	ARM64_SYS_REG(3, 3, 14, 8, 7),
+	ARM64_SYS_REG(3, 3, 14, 9, 0),
+	ARM64_SYS_REG(3, 3, 14, 9, 1),
+	ARM64_SYS_REG(3, 3, 14, 9, 2),
+	ARM64_SYS_REG(3, 3, 14, 9, 3),
+	ARM64_SYS_REG(3, 3, 14, 9, 4),
+	ARM64_SYS_REG(3, 3, 14, 9, 5),
+	ARM64_SYS_REG(3, 3, 14, 9, 6),
+	ARM64_SYS_REG(3, 3, 14, 9, 7),
+	ARM64_SYS_REG(3, 3, 14, 10, 0),
+	ARM64_SYS_REG(3, 3, 14, 10, 1),
+	ARM64_SYS_REG(3, 3, 14, 10, 2),
+	ARM64_SYS_REG(3, 3, 14, 10, 3),
+	ARM64_SYS_REG(3, 3, 14, 10, 4),
+	ARM64_SYS_REG(3, 3, 14, 10, 5),
+	ARM64_SYS_REG(3, 3, 14, 10, 6),
+	ARM64_SYS_REG(3, 3, 14, 10, 7),
+	ARM64_SYS_REG(3, 3, 14, 11, 0),
+	ARM64_SYS_REG(3, 3, 14, 11, 1),
+	ARM64_SYS_REG(3, 3, 14, 11, 2),
+	ARM64_SYS_REG(3, 3, 14, 11, 3),
+	ARM64_SYS_REG(3, 3, 14, 11, 4),
+	ARM64_SYS_REG(3, 3, 14, 11, 5),
+	ARM64_SYS_REG(3, 3, 14, 11, 6),
+	ARM64_SYS_REG(3, 3, 14, 12, 0),
+	ARM64_SYS_REG(3, 3, 14, 12, 1),
+	ARM64_SYS_REG(3, 3, 14, 12, 2),
+	ARM64_SYS_REG(3, 3, 14, 12, 3),
+	ARM64_SYS_REG(3, 3, 14, 12, 4),
+	ARM64_SYS_REG(3, 3, 14, 12, 5),
+	ARM64_SYS_REG(3, 3, 14, 12, 6),
+	ARM64_SYS_REG(3, 3, 14, 12, 7),
+	ARM64_SYS_REG(3, 3, 14, 13, 0),
+	ARM64_SYS_REG(3, 3, 14, 13, 1),
+	ARM64_SYS_REG(3, 3, 14, 13, 2),
+	ARM64_SYS_REG(3, 3, 14, 13, 3),
+	ARM64_SYS_REG(3, 3, 14, 13, 4),
+	ARM64_SYS_REG(3, 3, 14, 13, 5),
+	ARM64_SYS_REG(3, 3, 14, 13, 6),
+	ARM64_SYS_REG(3, 3, 14, 13, 7),
+	ARM64_SYS_REG(3, 3, 14, 14, 0),
+	ARM64_SYS_REG(3, 3, 14, 14, 1),
+	ARM64_SYS_REG(3, 3, 14, 14, 2),
+	ARM64_SYS_REG(3, 3, 14, 14, 3),
+	ARM64_SYS_REG(3, 3, 14, 14, 4),
+	ARM64_SYS_REG(3, 3, 14, 14, 5),
+	ARM64_SYS_REG(3, 3, 14, 14, 6),
+	ARM64_SYS_REG(3, 3, 14, 14, 7),
+	ARM64_SYS_REG(3, 3, 14, 15, 0),
+	ARM64_SYS_REG(3, 3, 14, 15, 1),
+	ARM64_SYS_REG(3, 3, 14, 15, 2),
+	ARM64_SYS_REG(3, 3, 14, 15, 3),
+	ARM64_SYS_REG(3, 3, 14, 15, 4),
+	ARM64_SYS_REG(3, 3, 14, 15, 5),
+	ARM64_SYS_REG(3, 3, 14, 15, 6),
+	ARM64_SYS_REG(3, 3, 14, 15, 7),	/* PMCCFILTR_EL0 */
+};
+
+static __u64 vregs[] = {
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[0]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[1]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[2]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[3]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[4]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[5]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[6]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[7]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[8]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[9]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[10]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[11]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[12]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[13]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[14]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[15]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[16]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[17]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[18]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[19]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[20]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[21]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[22]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[23]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[24]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[25]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[26]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[27]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[28]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[29]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[30]),
+	KVM_REG_ARM64 | KVM_REG_SIZE_U128 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.vregs[31]),
+};
+
+static __u64 sve_regs[] = {
+	KVM_REG_ARM64_SVE_VLS,
+	KVM_REG_ARM64_SVE_ZREG(0, 0),
+	KVM_REG_ARM64_SVE_ZREG(1, 0),
+	KVM_REG_ARM64_SVE_ZREG(2, 0),
+	KVM_REG_ARM64_SVE_ZREG(3, 0),
+	KVM_REG_ARM64_SVE_ZREG(4, 0),
+	KVM_REG_ARM64_SVE_ZREG(5, 0),
+	KVM_REG_ARM64_SVE_ZREG(6, 0),
+	KVM_REG_ARM64_SVE_ZREG(7, 0),
+	KVM_REG_ARM64_SVE_ZREG(8, 0),
+	KVM_REG_ARM64_SVE_ZREG(9, 0),
+	KVM_REG_ARM64_SVE_ZREG(10, 0),
+	KVM_REG_ARM64_SVE_ZREG(11, 0),
+	KVM_REG_ARM64_SVE_ZREG(12, 0),
+	KVM_REG_ARM64_SVE_ZREG(13, 0),
+	KVM_REG_ARM64_SVE_ZREG(14, 0),
+	KVM_REG_ARM64_SVE_ZREG(15, 0),
+	KVM_REG_ARM64_SVE_ZREG(16, 0),
+	KVM_REG_ARM64_SVE_ZREG(17, 0),
+	KVM_REG_ARM64_SVE_ZREG(18, 0),
+	KVM_REG_ARM64_SVE_ZREG(19, 0),
+	KVM_REG_ARM64_SVE_ZREG(20, 0),
+	KVM_REG_ARM64_SVE_ZREG(21, 0),
+	KVM_REG_ARM64_SVE_ZREG(22, 0),
+	KVM_REG_ARM64_SVE_ZREG(23, 0),
+	KVM_REG_ARM64_SVE_ZREG(24, 0),
+	KVM_REG_ARM64_SVE_ZREG(25, 0),
+	KVM_REG_ARM64_SVE_ZREG(26, 0),
+	KVM_REG_ARM64_SVE_ZREG(27, 0),
+	KVM_REG_ARM64_SVE_ZREG(28, 0),
+	KVM_REG_ARM64_SVE_ZREG(29, 0),
+	KVM_REG_ARM64_SVE_ZREG(30, 0),
+	KVM_REG_ARM64_SVE_ZREG(31, 0),
+	KVM_REG_ARM64_SVE_PREG(0, 0),
+	KVM_REG_ARM64_SVE_PREG(1, 0),
+	KVM_REG_ARM64_SVE_PREG(2, 0),
+	KVM_REG_ARM64_SVE_PREG(3, 0),
+	KVM_REG_ARM64_SVE_PREG(4, 0),
+	KVM_REG_ARM64_SVE_PREG(5, 0),
+	KVM_REG_ARM64_SVE_PREG(6, 0),
+	KVM_REG_ARM64_SVE_PREG(7, 0),
+	KVM_REG_ARM64_SVE_PREG(8, 0),
+	KVM_REG_ARM64_SVE_PREG(9, 0),
+	KVM_REG_ARM64_SVE_PREG(10, 0),
+	KVM_REG_ARM64_SVE_PREG(11, 0),
+	KVM_REG_ARM64_SVE_PREG(12, 0),
+	KVM_REG_ARM64_SVE_PREG(13, 0),
+	KVM_REG_ARM64_SVE_PREG(14, 0),
+	KVM_REG_ARM64_SVE_PREG(15, 0),
+	KVM_REG_ARM64_SVE_FFR(0),
+	ARM64_SYS_REG(3, 0, 1, 2, 0),   /* ZCR_EL1 */
+};
+
+static __u64 sve_rejects_set[] = {
+	KVM_REG_ARM64_SVE_VLS,
+};
+
+static __u64 pauth_addr_regs[] = {
+	ARM64_SYS_REG(3, 0, 2, 1, 0),	/* APIAKEYLO_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 1, 1),	/* APIAKEYHI_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 1, 2),	/* APIBKEYLO_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 1, 3),	/* APIBKEYHI_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 2, 0),	/* APDAKEYLO_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 2, 1),	/* APDAKEYHI_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 2, 2),	/* APDBKEYLO_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 2, 3)	/* APDBKEYHI_EL1 */
+};
+
+static __u64 pauth_generic_regs[] = {
+	ARM64_SYS_REG(3, 0, 2, 3, 0),	/* APGAKEYLO_EL1 */
+	ARM64_SYS_REG(3, 0, 2, 3, 1),	/* APGAKEYHI_EL1 */
+};
+
+#define BASE_SUBLIST \
+	{ "base", .regs = base_regs, .regs_n = ARRAY_SIZE(base_regs), }
+#define VREGS_SUBLIST \
+	{ "vregs", .regs = vregs, .regs_n = ARRAY_SIZE(vregs), }
+#define PMU_SUBLIST \
+	{ "pmu", .capability = KVM_CAP_ARM_PMU_V3, .feature = KVM_ARM_VCPU_PMU_V3, \
+	  .regs = pmu_regs, .regs_n = ARRAY_SIZE(pmu_regs), }
+#define SVE_SUBLIST \
+	{ "sve", .capability = KVM_CAP_ARM_SVE, .feature = KVM_ARM_VCPU_SVE, .finalize = true, \
+	  .regs = sve_regs, .regs_n = ARRAY_SIZE(sve_regs), \
+	  .rejects_set = sve_rejects_set, .rejects_set_n = ARRAY_SIZE(sve_rejects_set), }
+#define PAUTH_SUBLIST							\
+	{								\
+		.name 		= "pauth_address",			\
+		.capability	= KVM_CAP_ARM_PTRAUTH_ADDRESS,		\
+		.feature	= KVM_ARM_VCPU_PTRAUTH_ADDRESS,		\
+		.regs		= pauth_addr_regs,			\
+		.regs_n		= ARRAY_SIZE(pauth_addr_regs),		\
+	},								\
+	{								\
+		.name 		= "pauth_generic",			\
+		.capability	= KVM_CAP_ARM_PTRAUTH_GENERIC,		\
+		.feature	= KVM_ARM_VCPU_PTRAUTH_GENERIC,		\
+		.regs		= pauth_generic_regs,			\
+		.regs_n		= ARRAY_SIZE(pauth_generic_regs),	\
+	}
+
+static struct vcpu_reg_list vregs_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	VREGS_SUBLIST,
+	{0},
+	},
+};
+static struct vcpu_reg_list vregs_pmu_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	VREGS_SUBLIST,
+	PMU_SUBLIST,
+	{0},
+	},
+};
+static struct vcpu_reg_list sve_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	SVE_SUBLIST,
+	{0},
+	},
+};
+static struct vcpu_reg_list sve_pmu_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	SVE_SUBLIST,
+	PMU_SUBLIST,
+	{0},
+	},
+};
+static struct vcpu_reg_list pauth_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	VREGS_SUBLIST,
+	PAUTH_SUBLIST,
+	{0},
+	},
+};
+static struct vcpu_reg_list pauth_pmu_config = {
+	.sublists = {
+	BASE_SUBLIST,
+	VREGS_SUBLIST,
+	PAUTH_SUBLIST,
+	PMU_SUBLIST,
+	{0},
+	},
+};
+
+struct vcpu_reg_list *vcpu_configs[] = {
+	&vregs_config,
+	&vregs_pmu_config,
+	&sve_config,
+	&sve_pmu_config,
+	&pauth_config,
+	&pauth_pmu_config,
+};
+int vcpu_configs_n = ARRAY_SIZE(vcpu_configs);
diff --git a/tools/testing/selftests/kvm/aarch64/hypercalls.c b/tools/testing/selftests/kvm/aarch64/hypercalls.c
new file mode 100644
index 0000000000..31f66ba972
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/hypercalls.c
@@ -0,0 +1,308 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+/* hypercalls: Check the ARM64's psuedo-firmware bitmap register interface.
+ *
+ * The test validates the basic hypercall functionalities that are exposed
+ * via the psuedo-firmware bitmap register. This includes the registers'
+ * read/write behavior before and after the VM has started, and if the
+ * hypercalls are properly masked or unmasked to the guest when disabled or
+ * enabled from the KVM userspace, respectively.
+ */
+#include <errno.h>
+#include <linux/arm-smccc.h>
+#include <asm/kvm.h>
+#include <kvm_util.h>
+
+#include "processor.h"
+
+#define FW_REG_ULIMIT_VAL(max_feat_bit) (GENMASK(max_feat_bit, 0))
+
+/* Last valid bits of the bitmapped firmware registers */
+#define KVM_REG_ARM_STD_BMAP_BIT_MAX		0
+#define KVM_REG_ARM_STD_HYP_BMAP_BIT_MAX	0
+#define KVM_REG_ARM_VENDOR_HYP_BMAP_BIT_MAX	1
+
+struct kvm_fw_reg_info {
+	uint64_t reg;		/* Register definition */
+	uint64_t max_feat_bit;	/* Bit that represents the upper limit of the feature-map */
+};
+
+#define FW_REG_INFO(r)			\
+	{					\
+		.reg = r,			\
+		.max_feat_bit = r##_BIT_MAX,	\
+	}
+
+static const struct kvm_fw_reg_info fw_reg_info[] = {
+	FW_REG_INFO(KVM_REG_ARM_STD_BMAP),
+	FW_REG_INFO(KVM_REG_ARM_STD_HYP_BMAP),
+	FW_REG_INFO(KVM_REG_ARM_VENDOR_HYP_BMAP),
+};
+
+enum test_stage {
+	TEST_STAGE_REG_IFACE,
+	TEST_STAGE_HVC_IFACE_FEAT_DISABLED,
+	TEST_STAGE_HVC_IFACE_FEAT_ENABLED,
+	TEST_STAGE_HVC_IFACE_FALSE_INFO,
+	TEST_STAGE_END,
+};
+
+static int stage = TEST_STAGE_REG_IFACE;
+
+struct test_hvc_info {
+	uint32_t func_id;
+	uint64_t arg1;
+};
+
+#define TEST_HVC_INFO(f, a1)	\
+	{			\
+		.func_id = f,	\
+		.arg1 = a1,	\
+	}
+
+static const struct test_hvc_info hvc_info[] = {
+	/* KVM_REG_ARM_STD_BMAP */
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_VERSION, 0),
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_FEATURES, ARM_SMCCC_TRNG_RND64),
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_GET_UUID, 0),
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_RND32, 0),
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_RND64, 0),
+
+	/* KVM_REG_ARM_STD_HYP_BMAP */
+	TEST_HVC_INFO(ARM_SMCCC_ARCH_FEATURES_FUNC_ID, ARM_SMCCC_HV_PV_TIME_FEATURES),
+	TEST_HVC_INFO(ARM_SMCCC_HV_PV_TIME_FEATURES, ARM_SMCCC_HV_PV_TIME_ST),
+	TEST_HVC_INFO(ARM_SMCCC_HV_PV_TIME_ST, 0),
+
+	/* KVM_REG_ARM_VENDOR_HYP_BMAP */
+	TEST_HVC_INFO(ARM_SMCCC_VENDOR_HYP_KVM_FEATURES_FUNC_ID,
+			ARM_SMCCC_VENDOR_HYP_KVM_PTP_FUNC_ID),
+	TEST_HVC_INFO(ARM_SMCCC_VENDOR_HYP_CALL_UID_FUNC_ID, 0),
+	TEST_HVC_INFO(ARM_SMCCC_VENDOR_HYP_KVM_PTP_FUNC_ID, KVM_PTP_VIRT_COUNTER),
+};
+
+/* Feed false hypercall info to test the KVM behavior */
+static const struct test_hvc_info false_hvc_info[] = {
+	/* Feature support check against a different family of hypercalls */
+	TEST_HVC_INFO(ARM_SMCCC_TRNG_FEATURES, ARM_SMCCC_VENDOR_HYP_KVM_PTP_FUNC_ID),
+	TEST_HVC_INFO(ARM_SMCCC_ARCH_FEATURES_FUNC_ID, ARM_SMCCC_TRNG_RND64),
+	TEST_HVC_INFO(ARM_SMCCC_HV_PV_TIME_FEATURES, ARM_SMCCC_TRNG_RND64),
+};
+
+static void guest_test_hvc(const struct test_hvc_info *hc_info)
+{
+	unsigned int i;
+	struct arm_smccc_res res;
+	unsigned int hvc_info_arr_sz;
+
+	hvc_info_arr_sz =
+	hc_info == hvc_info ? ARRAY_SIZE(hvc_info) : ARRAY_SIZE(false_hvc_info);
+
+	for (i = 0; i < hvc_info_arr_sz; i++, hc_info++) {
+		memset(&res, 0, sizeof(res));
+		smccc_hvc(hc_info->func_id, hc_info->arg1, 0, 0, 0, 0, 0, 0, &res);
+
+		switch (stage) {
+		case TEST_STAGE_HVC_IFACE_FEAT_DISABLED:
+		case TEST_STAGE_HVC_IFACE_FALSE_INFO:
+			__GUEST_ASSERT(res.a0 == SMCCC_RET_NOT_SUPPORTED,
+				       "a0 = 0x%lx, func_id = 0x%x, arg1 = 0x%llx, stage = %u",
+					res.a0, hc_info->func_id, hc_info->arg1, stage);
+			break;
+		case TEST_STAGE_HVC_IFACE_FEAT_ENABLED:
+			__GUEST_ASSERT(res.a0 != SMCCC_RET_NOT_SUPPORTED,
+				       "a0 = 0x%lx, func_id = 0x%x, arg1 = 0x%llx, stage = %u",
+					res.a0, hc_info->func_id, hc_info->arg1, stage);
+			break;
+		default:
+			GUEST_FAIL("Unexpected stage = %u", stage);
+		}
+	}
+}
+
+static void guest_code(void)
+{
+	while (stage != TEST_STAGE_END) {
+		switch (stage) {
+		case TEST_STAGE_REG_IFACE:
+			break;
+		case TEST_STAGE_HVC_IFACE_FEAT_DISABLED:
+		case TEST_STAGE_HVC_IFACE_FEAT_ENABLED:
+			guest_test_hvc(hvc_info);
+			break;
+		case TEST_STAGE_HVC_IFACE_FALSE_INFO:
+			guest_test_hvc(false_hvc_info);
+			break;
+		default:
+			GUEST_FAIL("Unexpected stage = %u", stage);
+		}
+
+		GUEST_SYNC(stage);
+	}
+
+	GUEST_DONE();
+}
+
+struct st_time {
+	uint32_t rev;
+	uint32_t attr;
+	uint64_t st_time;
+};
+
+#define STEAL_TIME_SIZE		((sizeof(struct st_time) + 63) & ~63)
+#define ST_GPA_BASE		(1 << 30)
+
+static void steal_time_init(struct kvm_vcpu *vcpu)
+{
+	uint64_t st_ipa = (ulong)ST_GPA_BASE;
+	unsigned int gpages;
+
+	gpages = vm_calc_num_guest_pages(VM_MODE_DEFAULT, STEAL_TIME_SIZE);
+	vm_userspace_mem_region_add(vcpu->vm, VM_MEM_SRC_ANONYMOUS, ST_GPA_BASE, 1, gpages, 0);
+
+	vcpu_device_attr_set(vcpu, KVM_ARM_VCPU_PVTIME_CTRL,
+			     KVM_ARM_VCPU_PVTIME_IPA, &st_ipa);
+}
+
+static void test_fw_regs_before_vm_start(struct kvm_vcpu *vcpu)
+{
+	uint64_t val;
+	unsigned int i;
+	int ret;
+
+	for (i = 0; i < ARRAY_SIZE(fw_reg_info); i++) {
+		const struct kvm_fw_reg_info *reg_info = &fw_reg_info[i];
+
+		/* First 'read' should be an upper limit of the features supported */
+		vcpu_get_reg(vcpu, reg_info->reg, &val);
+		TEST_ASSERT(val == FW_REG_ULIMIT_VAL(reg_info->max_feat_bit),
+			"Expected all the features to be set for reg: 0x%lx; expected: 0x%lx; read: 0x%lx\n",
+			reg_info->reg, FW_REG_ULIMIT_VAL(reg_info->max_feat_bit), val);
+
+		/* Test a 'write' by disabling all the features of the register map */
+		ret = __vcpu_set_reg(vcpu, reg_info->reg, 0);
+		TEST_ASSERT(ret == 0,
+			"Failed to clear all the features of reg: 0x%lx; ret: %d\n",
+			reg_info->reg, errno);
+
+		vcpu_get_reg(vcpu, reg_info->reg, &val);
+		TEST_ASSERT(val == 0,
+			"Expected all the features to be cleared for reg: 0x%lx\n", reg_info->reg);
+
+		/*
+		 * Test enabling a feature that's not supported.
+		 * Avoid this check if all the bits are occupied.
+		 */
+		if (reg_info->max_feat_bit < 63) {
+			ret = __vcpu_set_reg(vcpu, reg_info->reg, BIT(reg_info->max_feat_bit + 1));
+			TEST_ASSERT(ret != 0 && errno == EINVAL,
+			"Unexpected behavior or return value (%d) while setting an unsupported feature for reg: 0x%lx\n",
+			errno, reg_info->reg);
+		}
+	}
+}
+
+static void test_fw_regs_after_vm_start(struct kvm_vcpu *vcpu)
+{
+	uint64_t val;
+	unsigned int i;
+	int ret;
+
+	for (i = 0; i < ARRAY_SIZE(fw_reg_info); i++) {
+		const struct kvm_fw_reg_info *reg_info = &fw_reg_info[i];
+
+		/*
+		 * Before starting the VM, the test clears all the bits.
+		 * Check if that's still the case.
+		 */
+		vcpu_get_reg(vcpu, reg_info->reg, &val);
+		TEST_ASSERT(val == 0,
+			"Expected all the features to be cleared for reg: 0x%lx\n",
+			reg_info->reg);
+
+		/*
+		 * Since the VM has run at least once, KVM shouldn't allow modification of
+		 * the registers and should return EBUSY. Set the registers and check for
+		 * the expected errno.
+		 */
+		ret = __vcpu_set_reg(vcpu, reg_info->reg, FW_REG_ULIMIT_VAL(reg_info->max_feat_bit));
+		TEST_ASSERT(ret != 0 && errno == EBUSY,
+		"Unexpected behavior or return value (%d) while setting a feature while VM is running for reg: 0x%lx\n",
+		errno, reg_info->reg);
+	}
+}
+
+static struct kvm_vm *test_vm_create(struct kvm_vcpu **vcpu)
+{
+	struct kvm_vm *vm;
+
+	vm = vm_create_with_one_vcpu(vcpu, guest_code);
+
+	steal_time_init(*vcpu);
+
+	return vm;
+}
+
+static void test_guest_stage(struct kvm_vm **vm, struct kvm_vcpu **vcpu)
+{
+	int prev_stage = stage;
+
+	pr_debug("Stage: %d\n", prev_stage);
+
+	/* Sync the stage early, the VM might be freed below. */
+	stage++;
+	sync_global_to_guest(*vm, stage);
+
+	switch (prev_stage) {
+	case TEST_STAGE_REG_IFACE:
+		test_fw_regs_after_vm_start(*vcpu);
+		break;
+	case TEST_STAGE_HVC_IFACE_FEAT_DISABLED:
+		/* Start a new VM so that all the features are now enabled by default */
+		kvm_vm_free(*vm);
+		*vm = test_vm_create(vcpu);
+		break;
+	case TEST_STAGE_HVC_IFACE_FEAT_ENABLED:
+	case TEST_STAGE_HVC_IFACE_FALSE_INFO:
+		break;
+	default:
+		TEST_FAIL("Unknown test stage: %d\n", prev_stage);
+	}
+}
+
+static void test_run(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	struct ucall uc;
+	bool guest_done = false;
+
+	vm = test_vm_create(&vcpu);
+
+	test_fw_regs_before_vm_start(vcpu);
+
+	while (!guest_done) {
+		vcpu_run(vcpu);
+
+		switch (get_ucall(vcpu, &uc)) {
+		case UCALL_SYNC:
+			test_guest_stage(&vm, &vcpu);
+			break;
+		case UCALL_DONE:
+			guest_done = true;
+			break;
+		case UCALL_ABORT:
+			REPORT_GUEST_ASSERT(uc);
+			break;
+		default:
+			TEST_FAIL("Unexpected guest exit\n");
+		}
+	}
+
+	kvm_vm_free(vm);
+}
+
+int main(void)
+{
+	test_run();
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/page_fault_test.c b/tools/testing/selftests/kvm/aarch64/page_fault_test.c
new file mode 100644
index 0000000000..47bb914ab2
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/page_fault_test.c
@@ -0,0 +1,1133 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * page_fault_test.c - Test stage 2 faults.
+ *
+ * This test tries different combinations of guest accesses (e.g., write,
+ * S1PTW), backing source type (e.g., anon) and types of faults (e.g., read on
+ * hugetlbfs with a hole). It checks that the expected handling method is
+ * called (e.g., uffd faults with the right address and write/read flag).
+ */
+#define _GNU_SOURCE
+#include <linux/bitmap.h>
+#include <fcntl.h>
+#include <test_util.h>
+#include <kvm_util.h>
+#include <processor.h>
+#include <asm/sysreg.h>
+#include <linux/bitfield.h>
+#include "guest_modes.h"
+#include "userfaultfd_util.h"
+
+/* Guest virtual addresses that point to the test page and its PTE. */
+#define TEST_GVA				0xc0000000
+#define TEST_EXEC_GVA				(TEST_GVA + 0x8)
+#define TEST_PTE_GVA				0xb0000000
+#define TEST_DATA				0x0123456789ABCDEF
+
+static uint64_t *guest_test_memory = (uint64_t *)TEST_GVA;
+
+#define CMD_NONE				(0)
+#define CMD_SKIP_TEST				(1ULL << 1)
+#define CMD_HOLE_PT				(1ULL << 2)
+#define CMD_HOLE_DATA				(1ULL << 3)
+#define CMD_CHECK_WRITE_IN_DIRTY_LOG		(1ULL << 4)
+#define CMD_CHECK_S1PTW_WR_IN_DIRTY_LOG		(1ULL << 5)
+#define CMD_CHECK_NO_WRITE_IN_DIRTY_LOG		(1ULL << 6)
+#define CMD_CHECK_NO_S1PTW_WR_IN_DIRTY_LOG	(1ULL << 7)
+#define CMD_SET_PTE_AF				(1ULL << 8)
+
+#define PREPARE_FN_NR				10
+#define CHECK_FN_NR				10
+
+static struct event_cnt {
+	int mmio_exits;
+	int fail_vcpu_runs;
+	int uffd_faults;
+	/* uffd_faults is incremented from multiple threads. */
+	pthread_mutex_t uffd_faults_mutex;
+} events;
+
+struct test_desc {
+	const char *name;
+	uint64_t mem_mark_cmd;
+	/* Skip the test if any prepare function returns false */
+	bool (*guest_prepare[PREPARE_FN_NR])(void);
+	void (*guest_test)(void);
+	void (*guest_test_check[CHECK_FN_NR])(void);
+	uffd_handler_t uffd_pt_handler;
+	uffd_handler_t uffd_data_handler;
+	void (*dabt_handler)(struct ex_regs *regs);
+	void (*iabt_handler)(struct ex_regs *regs);
+	void (*mmio_handler)(struct kvm_vm *vm, struct kvm_run *run);
+	void (*fail_vcpu_run_handler)(int ret);
+	uint32_t pt_memslot_flags;
+	uint32_t data_memslot_flags;
+	bool skip;
+	struct event_cnt expected_events;
+};
+
+struct test_params {
+	enum vm_mem_backing_src_type src_type;
+	struct test_desc *test_desc;
+};
+
+static inline void flush_tlb_page(uint64_t vaddr)
+{
+	uint64_t page = vaddr >> 12;
+
+	dsb(ishst);
+	asm volatile("tlbi vaae1is, %0" :: "r" (page));
+	dsb(ish);
+	isb();
+}
+
+static void guest_write64(void)
+{
+	uint64_t val;
+
+	WRITE_ONCE(*guest_test_memory, TEST_DATA);
+	val = READ_ONCE(*guest_test_memory);
+	GUEST_ASSERT_EQ(val, TEST_DATA);
+}
+
+/* Check the system for atomic instructions. */
+static bool guest_check_lse(void)
+{
+	uint64_t isar0 = read_sysreg(id_aa64isar0_el1);
+	uint64_t atomic;
+
+	atomic = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64ISAR0_ATOMICS), isar0);
+	return atomic >= 2;
+}
+
+static bool guest_check_dc_zva(void)
+{
+	uint64_t dczid = read_sysreg(dczid_el0);
+	uint64_t dzp = FIELD_GET(ARM64_FEATURE_MASK(DCZID_DZP), dczid);
+
+	return dzp == 0;
+}
+
+/* Compare and swap instruction. */
+static void guest_cas(void)
+{
+	uint64_t val;
+
+	GUEST_ASSERT(guest_check_lse());
+	asm volatile(".arch_extension lse\n"
+		     "casal %0, %1, [%2]\n"
+		     :: "r" (0ul), "r" (TEST_DATA), "r" (guest_test_memory));
+	val = READ_ONCE(*guest_test_memory);
+	GUEST_ASSERT_EQ(val, TEST_DATA);
+}
+
+static void guest_read64(void)
+{
+	uint64_t val;
+
+	val = READ_ONCE(*guest_test_memory);
+	GUEST_ASSERT_EQ(val, 0);
+}
+
+/* Address translation instruction */
+static void guest_at(void)
+{
+	uint64_t par;
+
+	asm volatile("at s1e1r, %0" :: "r" (guest_test_memory));
+	par = read_sysreg(par_el1);
+	isb();
+
+	/* Bit 1 indicates whether the AT was successful */
+	GUEST_ASSERT_EQ(par & 1, 0);
+}
+
+/*
+ * The size of the block written by "dc zva" is guaranteed to be between (2 <<
+ * 0) and (2 << 9), which is safe in our case as we need the write to happen
+ * for at least a word, and not more than a page.
+ */
+static void guest_dc_zva(void)
+{
+	uint16_t val;
+
+	asm volatile("dc zva, %0" :: "r" (guest_test_memory));
+	dsb(ish);
+	val = READ_ONCE(*guest_test_memory);
+	GUEST_ASSERT_EQ(val, 0);
+}
+
+/*
+ * Pre-indexing loads and stores don't have a valid syndrome (ESR_EL2.ISV==0).
+ * And that's special because KVM must take special care with those: they
+ * should still count as accesses for dirty logging or user-faulting, but
+ * should be handled differently on mmio.
+ */
+static void guest_ld_preidx(void)
+{
+	uint64_t val;
+	uint64_t addr = TEST_GVA - 8;
+
+	/*
+	 * This ends up accessing "TEST_GVA + 8 - 8", where "TEST_GVA - 8" is
+	 * in a gap between memslots not backing by anything.
+	 */
+	asm volatile("ldr %0, [%1, #8]!"
+		     : "=r" (val), "+r" (addr));
+	GUEST_ASSERT_EQ(val, 0);
+	GUEST_ASSERT_EQ(addr, TEST_GVA);
+}
+
+static void guest_st_preidx(void)
+{
+	uint64_t val = TEST_DATA;
+	uint64_t addr = TEST_GVA - 8;
+
+	asm volatile("str %0, [%1, #8]!"
+		     : "+r" (val), "+r" (addr));
+
+	GUEST_ASSERT_EQ(addr, TEST_GVA);
+	val = READ_ONCE(*guest_test_memory);
+}
+
+static bool guest_set_ha(void)
+{
+	uint64_t mmfr1 = read_sysreg(id_aa64mmfr1_el1);
+	uint64_t hadbs, tcr;
+
+	/* Skip if HA is not supported. */
+	hadbs = FIELD_GET(ARM64_FEATURE_MASK(ID_AA64MMFR1_HADBS), mmfr1);
+	if (hadbs == 0)
+		return false;
+
+	tcr = read_sysreg(tcr_el1) | TCR_EL1_HA;
+	write_sysreg(tcr, tcr_el1);
+	isb();
+
+	return true;
+}
+
+static bool guest_clear_pte_af(void)
+{
+	*((uint64_t *)TEST_PTE_GVA) &= ~PTE_AF;
+	flush_tlb_page(TEST_GVA);
+
+	return true;
+}
+
+static void guest_check_pte_af(void)
+{
+	dsb(ish);
+	GUEST_ASSERT_EQ(*((uint64_t *)TEST_PTE_GVA) & PTE_AF, PTE_AF);
+}
+
+static void guest_check_write_in_dirty_log(void)
+{
+	GUEST_SYNC(CMD_CHECK_WRITE_IN_DIRTY_LOG);
+}
+
+static void guest_check_no_write_in_dirty_log(void)
+{
+	GUEST_SYNC(CMD_CHECK_NO_WRITE_IN_DIRTY_LOG);
+}
+
+static void guest_check_s1ptw_wr_in_dirty_log(void)
+{
+	GUEST_SYNC(CMD_CHECK_S1PTW_WR_IN_DIRTY_LOG);
+}
+
+static void guest_check_no_s1ptw_wr_in_dirty_log(void)
+{
+	GUEST_SYNC(CMD_CHECK_NO_S1PTW_WR_IN_DIRTY_LOG);
+}
+
+static void guest_exec(void)
+{
+	int (*code)(void) = (int (*)(void))TEST_EXEC_GVA;
+	int ret;
+
+	ret = code();
+	GUEST_ASSERT_EQ(ret, 0x77);
+}
+
+static bool guest_prepare(struct test_desc *test)
+{
+	bool (*prepare_fn)(void);
+	int i;
+
+	for (i = 0; i < PREPARE_FN_NR; i++) {
+		prepare_fn = test->guest_prepare[i];
+		if (prepare_fn && !prepare_fn())
+			return false;
+	}
+
+	return true;
+}
+
+static void guest_test_check(struct test_desc *test)
+{
+	void (*check_fn)(void);
+	int i;
+
+	for (i = 0; i < CHECK_FN_NR; i++) {
+		check_fn = test->guest_test_check[i];
+		if (check_fn)
+			check_fn();
+	}
+}
+
+static void guest_code(struct test_desc *test)
+{
+	if (!guest_prepare(test))
+		GUEST_SYNC(CMD_SKIP_TEST);
+
+	GUEST_SYNC(test->mem_mark_cmd);
+
+	if (test->guest_test)
+		test->guest_test();
+
+	guest_test_check(test);
+	GUEST_DONE();
+}
+
+static void no_dabt_handler(struct ex_regs *regs)
+{
+	GUEST_FAIL("Unexpected dabt, far_el1 = 0x%llx", read_sysreg(far_el1));
+}
+
+static void no_iabt_handler(struct ex_regs *regs)
+{
+	GUEST_FAIL("Unexpected iabt, pc = 0x%lx", regs->pc);
+}
+
+static struct uffd_args {
+	char *copy;
+	void *hva;
+	uint64_t paging_size;
+} pt_args, data_args;
+
+/* Returns true to continue the test, and false if it should be skipped. */
+static int uffd_generic_handler(int uffd_mode, int uffd, struct uffd_msg *msg,
+				struct uffd_args *args)
+{
+	uint64_t addr = msg->arg.pagefault.address;
+	uint64_t flags = msg->arg.pagefault.flags;
+	struct uffdio_copy copy;
+	int ret;
+
+	TEST_ASSERT(uffd_mode == UFFDIO_REGISTER_MODE_MISSING,
+		    "The only expected UFFD mode is MISSING");
+	TEST_ASSERT_EQ(addr, (uint64_t)args->hva);
+
+	pr_debug("uffd fault: addr=%p write=%d\n",
+		 (void *)addr, !!(flags & UFFD_PAGEFAULT_FLAG_WRITE));
+
+	copy.src = (uint64_t)args->copy;
+	copy.dst = addr;
+	copy.len = args->paging_size;
+	copy.mode = 0;
+
+	ret = ioctl(uffd, UFFDIO_COPY, &copy);
+	if (ret == -1) {
+		pr_info("Failed UFFDIO_COPY in 0x%lx with errno: %d\n",
+			addr, errno);
+		return ret;
+	}
+
+	pthread_mutex_lock(&events.uffd_faults_mutex);
+	events.uffd_faults += 1;
+	pthread_mutex_unlock(&events.uffd_faults_mutex);
+	return 0;
+}
+
+static int uffd_pt_handler(int mode, int uffd, struct uffd_msg *msg)
+{
+	return uffd_generic_handler(mode, uffd, msg, &pt_args);
+}
+
+static int uffd_data_handler(int mode, int uffd, struct uffd_msg *msg)
+{
+	return uffd_generic_handler(mode, uffd, msg, &data_args);
+}
+
+static void setup_uffd_args(struct userspace_mem_region *region,
+			    struct uffd_args *args)
+{
+	args->hva = (void *)region->region.userspace_addr;
+	args->paging_size = region->region.memory_size;
+
+	args->copy = malloc(args->paging_size);
+	TEST_ASSERT(args->copy, "Failed to allocate data copy.");
+	memcpy(args->copy, args->hva, args->paging_size);
+}
+
+static void setup_uffd(struct kvm_vm *vm, struct test_params *p,
+		       struct uffd_desc **pt_uffd, struct uffd_desc **data_uffd)
+{
+	struct test_desc *test = p->test_desc;
+	int uffd_mode = UFFDIO_REGISTER_MODE_MISSING;
+
+	setup_uffd_args(vm_get_mem_region(vm, MEM_REGION_PT), &pt_args);
+	setup_uffd_args(vm_get_mem_region(vm, MEM_REGION_TEST_DATA), &data_args);
+
+	*pt_uffd = NULL;
+	if (test->uffd_pt_handler)
+		*pt_uffd = uffd_setup_demand_paging(uffd_mode, 0,
+						    pt_args.hva,
+						    pt_args.paging_size,
+						    test->uffd_pt_handler);
+
+	*data_uffd = NULL;
+	if (test->uffd_data_handler)
+		*data_uffd = uffd_setup_demand_paging(uffd_mode, 0,
+						      data_args.hva,
+						      data_args.paging_size,
+						      test->uffd_data_handler);
+}
+
+static void free_uffd(struct test_desc *test, struct uffd_desc *pt_uffd,
+		      struct uffd_desc *data_uffd)
+{
+	if (test->uffd_pt_handler)
+		uffd_stop_demand_paging(pt_uffd);
+	if (test->uffd_data_handler)
+		uffd_stop_demand_paging(data_uffd);
+
+	free(pt_args.copy);
+	free(data_args.copy);
+}
+
+static int uffd_no_handler(int mode, int uffd, struct uffd_msg *msg)
+{
+	TEST_FAIL("There was no UFFD fault expected.");
+	return -1;
+}
+
+/* Returns false if the test should be skipped. */
+static bool punch_hole_in_backing_store(struct kvm_vm *vm,
+					struct userspace_mem_region *region)
+{
+	void *hva = (void *)region->region.userspace_addr;
+	uint64_t paging_size = region->region.memory_size;
+	int ret, fd = region->fd;
+
+	if (fd != -1) {
+		ret = fallocate(fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE,
+				0, paging_size);
+		TEST_ASSERT(ret == 0, "fallocate failed\n");
+	} else {
+		ret = madvise(hva, paging_size, MADV_DONTNEED);
+		TEST_ASSERT(ret == 0, "madvise failed\n");
+	}
+
+	return true;
+}
+
+static void mmio_on_test_gpa_handler(struct kvm_vm *vm, struct kvm_run *run)
+{
+	struct userspace_mem_region *region;
+	void *hva;
+
+	region = vm_get_mem_region(vm, MEM_REGION_TEST_DATA);
+	hva = (void *)region->region.userspace_addr;
+
+	TEST_ASSERT_EQ(run->mmio.phys_addr, region->region.guest_phys_addr);
+
+	memcpy(hva, run->mmio.data, run->mmio.len);
+	events.mmio_exits += 1;
+}
+
+static void mmio_no_handler(struct kvm_vm *vm, struct kvm_run *run)
+{
+	uint64_t data;
+
+	memcpy(&data, run->mmio.data, sizeof(data));
+	pr_debug("addr=%lld len=%d w=%d data=%lx\n",
+		 run->mmio.phys_addr, run->mmio.len,
+		 run->mmio.is_write, data);
+	TEST_FAIL("There was no MMIO exit expected.");
+}
+
+static bool check_write_in_dirty_log(struct kvm_vm *vm,
+				     struct userspace_mem_region *region,
+				     uint64_t host_pg_nr)
+{
+	unsigned long *bmap;
+	bool first_page_dirty;
+	uint64_t size = region->region.memory_size;
+
+	/* getpage_size() is not always equal to vm->page_size */
+	bmap = bitmap_zalloc(size / getpagesize());
+	kvm_vm_get_dirty_log(vm, region->region.slot, bmap);
+	first_page_dirty = test_bit(host_pg_nr, bmap);
+	free(bmap);
+	return first_page_dirty;
+}
+
+/* Returns true to continue the test, and false if it should be skipped. */
+static bool handle_cmd(struct kvm_vm *vm, int cmd)
+{
+	struct userspace_mem_region *data_region, *pt_region;
+	bool continue_test = true;
+	uint64_t pte_gpa, pte_pg;
+
+	data_region = vm_get_mem_region(vm, MEM_REGION_TEST_DATA);
+	pt_region = vm_get_mem_region(vm, MEM_REGION_PT);
+	pte_gpa = addr_hva2gpa(vm, virt_get_pte_hva(vm, TEST_GVA));
+	pte_pg = (pte_gpa - pt_region->region.guest_phys_addr) / getpagesize();
+
+	if (cmd == CMD_SKIP_TEST)
+		continue_test = false;
+
+	if (cmd & CMD_HOLE_PT)
+		continue_test = punch_hole_in_backing_store(vm, pt_region);
+	if (cmd & CMD_HOLE_DATA)
+		continue_test = punch_hole_in_backing_store(vm, data_region);
+	if (cmd & CMD_CHECK_WRITE_IN_DIRTY_LOG)
+		TEST_ASSERT(check_write_in_dirty_log(vm, data_region, 0),
+			    "Missing write in dirty log");
+	if (cmd & CMD_CHECK_S1PTW_WR_IN_DIRTY_LOG)
+		TEST_ASSERT(check_write_in_dirty_log(vm, pt_region, pte_pg),
+			    "Missing s1ptw write in dirty log");
+	if (cmd & CMD_CHECK_NO_WRITE_IN_DIRTY_LOG)
+		TEST_ASSERT(!check_write_in_dirty_log(vm, data_region, 0),
+			    "Unexpected write in dirty log");
+	if (cmd & CMD_CHECK_NO_S1PTW_WR_IN_DIRTY_LOG)
+		TEST_ASSERT(!check_write_in_dirty_log(vm, pt_region, pte_pg),
+			    "Unexpected s1ptw write in dirty log");
+
+	return continue_test;
+}
+
+void fail_vcpu_run_no_handler(int ret)
+{
+	TEST_FAIL("Unexpected vcpu run failure\n");
+}
+
+void fail_vcpu_run_mmio_no_syndrome_handler(int ret)
+{
+	TEST_ASSERT(errno == ENOSYS,
+		    "The mmio handler should have returned not implemented.");
+	events.fail_vcpu_runs += 1;
+}
+
+typedef uint32_t aarch64_insn_t;
+extern aarch64_insn_t __exec_test[2];
+
+noinline void __return_0x77(void)
+{
+	asm volatile("__exec_test: mov x0, #0x77\n"
+		     "ret\n");
+}
+
+/*
+ * Note that this function runs on the host before the test VM starts: there's
+ * no need to sync the D$ and I$ caches.
+ */
+static void load_exec_code_for_test(struct kvm_vm *vm)
+{
+	uint64_t *code;
+	struct userspace_mem_region *region;
+	void *hva;
+
+	region = vm_get_mem_region(vm, MEM_REGION_TEST_DATA);
+	hva = (void *)region->region.userspace_addr;
+
+	assert(TEST_EXEC_GVA > TEST_GVA);
+	code = hva + TEST_EXEC_GVA - TEST_GVA;
+	memcpy(code, __exec_test, sizeof(__exec_test));
+}
+
+static void setup_abort_handlers(struct kvm_vm *vm, struct kvm_vcpu *vcpu,
+				 struct test_desc *test)
+{
+	vm_init_descriptor_tables(vm);
+	vcpu_init_descriptor_tables(vcpu);
+
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_DABT, no_dabt_handler);
+	vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT,
+				ESR_EC_IABT, no_iabt_handler);
+}
+
+static void setup_gva_maps(struct kvm_vm *vm)
+{
+	struct userspace_mem_region *region;
+	uint64_t pte_gpa;
+
+	region = vm_get_mem_region(vm, MEM_REGION_TEST_DATA);
+	/* Map TEST_GVA first. This will install a new PTE. */
+	virt_pg_map(vm, TEST_GVA, region->region.guest_phys_addr);
+	/* Then map TEST_PTE_GVA to the above PTE. */
+	pte_gpa = addr_hva2gpa(vm, virt_get_pte_hva(vm, TEST_GVA));
+	virt_pg_map(vm, TEST_PTE_GVA, pte_gpa);
+}
+
+enum pf_test_memslots {
+	CODE_AND_DATA_MEMSLOT,
+	PAGE_TABLE_MEMSLOT,
+	TEST_DATA_MEMSLOT,
+};
+
+/*
+ * Create a memslot for code and data at pfn=0, and test-data and PT ones
+ * at max_gfn.
+ */
+static void setup_memslots(struct kvm_vm *vm, struct test_params *p)
+{
+	uint64_t backing_src_pagesz = get_backing_src_pagesz(p->src_type);
+	uint64_t guest_page_size = vm->page_size;
+	uint64_t max_gfn = vm_compute_max_gfn(vm);
+	/* Enough for 2M of code when using 4K guest pages. */
+	uint64_t code_npages = 512;
+	uint64_t pt_size, data_size, data_gpa;
+
+	/*
+	 * This test requires 1 pgd, 2 pud, 4 pmd, and 6 pte pages when using
+	 * VM_MODE_P48V48_4K. Note that the .text takes ~1.6MBs.  That's 13
+	 * pages. VM_MODE_P48V48_4K is the mode with most PT pages; let's use
+	 * twice that just in case.
+	 */
+	pt_size = 26 * guest_page_size;
+
+	/* memslot sizes and gpa's must be aligned to the backing page size */
+	pt_size = align_up(pt_size, backing_src_pagesz);
+	data_size = align_up(guest_page_size, backing_src_pagesz);
+	data_gpa = (max_gfn * guest_page_size) - data_size;
+	data_gpa = align_down(data_gpa, backing_src_pagesz);
+
+	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS, 0,
+				    CODE_AND_DATA_MEMSLOT, code_npages, 0);
+	vm->memslots[MEM_REGION_CODE] = CODE_AND_DATA_MEMSLOT;
+	vm->memslots[MEM_REGION_DATA] = CODE_AND_DATA_MEMSLOT;
+
+	vm_userspace_mem_region_add(vm, p->src_type, data_gpa - pt_size,
+				    PAGE_TABLE_MEMSLOT, pt_size / guest_page_size,
+				    p->test_desc->pt_memslot_flags);
+	vm->memslots[MEM_REGION_PT] = PAGE_TABLE_MEMSLOT;
+
+	vm_userspace_mem_region_add(vm, p->src_type, data_gpa, TEST_DATA_MEMSLOT,
+				    data_size / guest_page_size,
+				    p->test_desc->data_memslot_flags);
+	vm->memslots[MEM_REGION_TEST_DATA] = TEST_DATA_MEMSLOT;
+}
+
+static void setup_ucall(struct kvm_vm *vm)
+{
+	struct userspace_mem_region *region = vm_get_mem_region(vm, MEM_REGION_TEST_DATA);
+
+	ucall_init(vm, region->region.guest_phys_addr + region->region.memory_size);
+}
+
+static void setup_default_handlers(struct test_desc *test)
+{
+	if (!test->mmio_handler)
+		test->mmio_handler = mmio_no_handler;
+
+	if (!test->fail_vcpu_run_handler)
+		test->fail_vcpu_run_handler = fail_vcpu_run_no_handler;
+}
+
+static void check_event_counts(struct test_desc *test)
+{
+	TEST_ASSERT_EQ(test->expected_events.uffd_faults, events.uffd_faults);
+	TEST_ASSERT_EQ(test->expected_events.mmio_exits, events.mmio_exits);
+	TEST_ASSERT_EQ(test->expected_events.fail_vcpu_runs, events.fail_vcpu_runs);
+}
+
+static void print_test_banner(enum vm_guest_mode mode, struct test_params *p)
+{
+	struct test_desc *test = p->test_desc;
+
+	pr_debug("Test: %s\n", test->name);
+	pr_debug("Testing guest mode: %s\n", vm_guest_mode_string(mode));
+	pr_debug("Testing memory backing src type: %s\n",
+		 vm_mem_backing_src_alias(p->src_type)->name);
+}
+
+static void reset_event_counts(void)
+{
+	memset(&events, 0, sizeof(events));
+}
+
+/*
+ * This function either succeeds, skips the test (after setting test->skip), or
+ * fails with a TEST_FAIL that aborts all tests.
+ */
+static void vcpu_run_loop(struct kvm_vm *vm, struct kvm_vcpu *vcpu,
+			  struct test_desc *test)
+{
+	struct kvm_run *run;
+	struct ucall uc;
+	int ret;
+
+	run = vcpu->run;
+
+	for (;;) {
+		ret = _vcpu_run(vcpu);
+		if (ret) {
+			test->fail_vcpu_run_handler(ret);
+			goto done;
+		}
+
+		switch (get_ucall(vcpu, &uc)) {
+		case UCALL_SYNC:
+			if (!handle_cmd(vm, uc.args[1])) {
+				test->skip = true;
+				goto done;
+			}
+			break;
+		case UCALL_ABORT:
+			REPORT_GUEST_ASSERT(uc);
+			break;
+		case UCALL_DONE:
+			goto done;
+		case UCALL_NONE:
+			if (run->exit_reason == KVM_EXIT_MMIO)
+				test->mmio_handler(vm, run);
+			break;
+		default:
+			TEST_FAIL("Unknown ucall %lu", uc.cmd);
+		}
+	}
+
+done:
+	pr_debug(test->skip ? "Skipped.\n" : "Done.\n");
+}
+
+static void run_test(enum vm_guest_mode mode, void *arg)
+{
+	struct test_params *p = (struct test_params *)arg;
+	struct test_desc *test = p->test_desc;
+	struct kvm_vm *vm;
+	struct kvm_vcpu *vcpu;
+	struct uffd_desc *pt_uffd, *data_uffd;
+
+	print_test_banner(mode, p);
+
+	vm = ____vm_create(mode);
+	setup_memslots(vm, p);
+	kvm_vm_elf_load(vm, program_invocation_name);
+	setup_ucall(vm);
+	vcpu = vm_vcpu_add(vm, 0, guest_code);
+
+	setup_gva_maps(vm);
+
+	reset_event_counts();
+
+	/*
+	 * Set some code in the data memslot for the guest to execute (only
+	 * applicable to the EXEC tests). This has to be done before
+	 * setup_uffd() as that function copies the memslot data for the uffd
+	 * handler.
+	 */
+	load_exec_code_for_test(vm);
+	setup_uffd(vm, p, &pt_uffd, &data_uffd);
+	setup_abort_handlers(vm, vcpu, test);
+	setup_default_handlers(test);
+	vcpu_args_set(vcpu, 1, test);
+
+	vcpu_run_loop(vm, vcpu, test);
+
+	kvm_vm_free(vm);
+	free_uffd(test, pt_uffd, data_uffd);
+
+	/*
+	 * Make sure we check the events after the uffd threads have exited,
+	 * which means they updated their respective event counters.
+	 */
+	if (!test->skip)
+		check_event_counts(test);
+}
+
+static void help(char *name)
+{
+	puts("");
+	printf("usage: %s [-h] [-s mem-type]\n", name);
+	puts("");
+	guest_modes_help();
+	backing_src_help("-s");
+	puts("");
+}
+
+#define SNAME(s)			#s
+#define SCAT2(a, b)			SNAME(a ## _ ## b)
+#define SCAT3(a, b, c)			SCAT2(a, SCAT2(b, c))
+#define SCAT4(a, b, c, d)		SCAT2(a, SCAT3(b, c, d))
+
+#define _CHECK(_test)			_CHECK_##_test
+#define _PREPARE(_test)			_PREPARE_##_test
+#define _PREPARE_guest_read64		NULL
+#define _PREPARE_guest_ld_preidx	NULL
+#define _PREPARE_guest_write64		NULL
+#define _PREPARE_guest_st_preidx	NULL
+#define _PREPARE_guest_exec		NULL
+#define _PREPARE_guest_at		NULL
+#define _PREPARE_guest_dc_zva		guest_check_dc_zva
+#define _PREPARE_guest_cas		guest_check_lse
+
+/* With or without access flag checks */
+#define _PREPARE_with_af		guest_set_ha, guest_clear_pte_af
+#define _PREPARE_no_af			NULL
+#define _CHECK_with_af			guest_check_pte_af
+#define _CHECK_no_af			NULL
+
+/* Performs an access and checks that no faults were triggered. */
+#define TEST_ACCESS(_access, _with_af, _mark_cmd)				\
+{										\
+	.name			= SCAT3(_access, _with_af, #_mark_cmd),		\
+	.guest_prepare		= { _PREPARE(_with_af),				\
+				    _PREPARE(_access) },			\
+	.mem_mark_cmd		= _mark_cmd,					\
+	.guest_test		= _access,					\
+	.guest_test_check	= { _CHECK(_with_af) },				\
+	.expected_events	= { 0 },					\
+}
+
+#define TEST_UFFD(_access, _with_af, _mark_cmd,					\
+		  _uffd_data_handler, _uffd_pt_handler, _uffd_faults)		\
+{										\
+	.name			= SCAT4(uffd, _access, _with_af, #_mark_cmd),	\
+	.guest_prepare		= { _PREPARE(_with_af),				\
+				    _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.mem_mark_cmd		= _mark_cmd,					\
+	.guest_test_check	= { _CHECK(_with_af) },				\
+	.uffd_data_handler	= _uffd_data_handler,				\
+	.uffd_pt_handler	= _uffd_pt_handler,				\
+	.expected_events	= { .uffd_faults = _uffd_faults, },		\
+}
+
+#define TEST_DIRTY_LOG(_access, _with_af, _test_check, _pt_check)		\
+{										\
+	.name			= SCAT3(dirty_log, _access, _with_af),		\
+	.data_memslot_flags	= KVM_MEM_LOG_DIRTY_PAGES,			\
+	.pt_memslot_flags	= KVM_MEM_LOG_DIRTY_PAGES,			\
+	.guest_prepare		= { _PREPARE(_with_af),				\
+				    _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.guest_test_check	= { _CHECK(_with_af), _test_check, _pt_check },	\
+	.expected_events	= { 0 },					\
+}
+
+#define TEST_UFFD_AND_DIRTY_LOG(_access, _with_af, _uffd_data_handler,		\
+				_uffd_faults, _test_check, _pt_check)		\
+{										\
+	.name			= SCAT3(uffd_and_dirty_log, _access, _with_af),	\
+	.data_memslot_flags	= KVM_MEM_LOG_DIRTY_PAGES,			\
+	.pt_memslot_flags	= KVM_MEM_LOG_DIRTY_PAGES,			\
+	.guest_prepare		= { _PREPARE(_with_af),				\
+				    _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.mem_mark_cmd		= CMD_HOLE_DATA | CMD_HOLE_PT,			\
+	.guest_test_check	= { _CHECK(_with_af), _test_check, _pt_check },	\
+	.uffd_data_handler	= _uffd_data_handler,				\
+	.uffd_pt_handler	= uffd_pt_handler,				\
+	.expected_events	= { .uffd_faults = _uffd_faults, },		\
+}
+
+#define TEST_RO_MEMSLOT(_access, _mmio_handler, _mmio_exits)			\
+{										\
+	.name			= SCAT2(ro_memslot, _access),			\
+	.data_memslot_flags	= KVM_MEM_READONLY,				\
+	.pt_memslot_flags	= KVM_MEM_READONLY,				\
+	.guest_prepare		= { _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.mmio_handler		= _mmio_handler,				\
+	.expected_events	= { .mmio_exits = _mmio_exits },		\
+}
+
+#define TEST_RO_MEMSLOT_NO_SYNDROME(_access)					\
+{										\
+	.name			= SCAT2(ro_memslot_no_syndrome, _access),	\
+	.data_memslot_flags	= KVM_MEM_READONLY,				\
+	.pt_memslot_flags	= KVM_MEM_READONLY,				\
+	.guest_test		= _access,					\
+	.fail_vcpu_run_handler	= fail_vcpu_run_mmio_no_syndrome_handler,	\
+	.expected_events	= { .fail_vcpu_runs = 1 },			\
+}
+
+#define TEST_RO_MEMSLOT_AND_DIRTY_LOG(_access, _mmio_handler, _mmio_exits,	\
+				      _test_check)				\
+{										\
+	.name			= SCAT2(ro_memslot, _access),			\
+	.data_memslot_flags	= KVM_MEM_READONLY | KVM_MEM_LOG_DIRTY_PAGES,	\
+	.pt_memslot_flags	= KVM_MEM_READONLY | KVM_MEM_LOG_DIRTY_PAGES,	\
+	.guest_prepare		= { _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.guest_test_check	= { _test_check },				\
+	.mmio_handler		= _mmio_handler,				\
+	.expected_events	= { .mmio_exits = _mmio_exits},			\
+}
+
+#define TEST_RO_MEMSLOT_NO_SYNDROME_AND_DIRTY_LOG(_access, _test_check)		\
+{										\
+	.name			= SCAT2(ro_memslot_no_syn_and_dlog, _access),	\
+	.data_memslot_flags	= KVM_MEM_READONLY | KVM_MEM_LOG_DIRTY_PAGES,	\
+	.pt_memslot_flags	= KVM_MEM_READONLY | KVM_MEM_LOG_DIRTY_PAGES,	\
+	.guest_test		= _access,					\
+	.guest_test_check	= { _test_check },				\
+	.fail_vcpu_run_handler	= fail_vcpu_run_mmio_no_syndrome_handler,	\
+	.expected_events	= { .fail_vcpu_runs = 1 },			\
+}
+
+#define TEST_RO_MEMSLOT_AND_UFFD(_access, _mmio_handler, _mmio_exits,		\
+				 _uffd_data_handler, _uffd_faults)		\
+{										\
+	.name			= SCAT2(ro_memslot_uffd, _access),		\
+	.data_memslot_flags	= KVM_MEM_READONLY,				\
+	.pt_memslot_flags	= KVM_MEM_READONLY,				\
+	.mem_mark_cmd		= CMD_HOLE_DATA | CMD_HOLE_PT,			\
+	.guest_prepare		= { _PREPARE(_access) },			\
+	.guest_test		= _access,					\
+	.uffd_data_handler	= _uffd_data_handler,				\
+	.uffd_pt_handler	= uffd_pt_handler,				\
+	.mmio_handler		= _mmio_handler,				\
+	.expected_events	= { .mmio_exits = _mmio_exits,			\
+				    .uffd_faults = _uffd_faults },		\
+}
+
+#define TEST_RO_MEMSLOT_NO_SYNDROME_AND_UFFD(_access, _uffd_data_handler,	\
+					     _uffd_faults)			\
+{										\
+	.name			= SCAT2(ro_memslot_no_syndrome, _access),	\
+	.data_memslot_flags	= KVM_MEM_READONLY,				\
+	.pt_memslot_flags	= KVM_MEM_READONLY,				\
+	.mem_mark_cmd		= CMD_HOLE_DATA | CMD_HOLE_PT,			\
+	.guest_test		= _access,					\
+	.uffd_data_handler	= _uffd_data_handler,				\
+	.uffd_pt_handler	= uffd_pt_handler,			\
+	.fail_vcpu_run_handler	= fail_vcpu_run_mmio_no_syndrome_handler,	\
+	.expected_events	= { .fail_vcpu_runs = 1,			\
+				    .uffd_faults = _uffd_faults },		\
+}
+
+static struct test_desc tests[] = {
+
+	/* Check that HW is setting the Access Flag (AF) (sanity checks). */
+	TEST_ACCESS(guest_read64, with_af, CMD_NONE),
+	TEST_ACCESS(guest_ld_preidx, with_af, CMD_NONE),
+	TEST_ACCESS(guest_cas, with_af, CMD_NONE),
+	TEST_ACCESS(guest_write64, with_af, CMD_NONE),
+	TEST_ACCESS(guest_st_preidx, with_af, CMD_NONE),
+	TEST_ACCESS(guest_dc_zva, with_af, CMD_NONE),
+	TEST_ACCESS(guest_exec, with_af, CMD_NONE),
+
+	/*
+	 * Punch a hole in the data backing store, and then try multiple
+	 * accesses: reads should rturn zeroes, and writes should
+	 * re-populate the page. Moreover, the test also check that no
+	 * exception was generated in the guest.  Note that this
+	 * reading/writing behavior is the same as reading/writing a
+	 * punched page (with fallocate(FALLOC_FL_PUNCH_HOLE)) from
+	 * userspace.
+	 */
+	TEST_ACCESS(guest_read64, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_cas, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_ld_preidx, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_write64, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_st_preidx, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_at, no_af, CMD_HOLE_DATA),
+	TEST_ACCESS(guest_dc_zva, no_af, CMD_HOLE_DATA),
+
+	/*
+	 * Punch holes in the data and PT backing stores and mark them for
+	 * userfaultfd handling. This should result in 2 faults: the access
+	 * on the data backing store, and its respective S1 page table walk
+	 * (S1PTW).
+	 */
+	TEST_UFFD(guest_read64, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_read64, no_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_cas, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	/*
+	 * Can't test guest_at with_af as it's IMPDEF whether the AF is set.
+	 * The S1PTW fault should still be marked as a write.
+	 */
+	TEST_UFFD(guest_at, no_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_no_handler, uffd_pt_handler, 1),
+	TEST_UFFD(guest_ld_preidx, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_write64, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_dc_zva, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_st_preidx, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+	TEST_UFFD(guest_exec, with_af, CMD_HOLE_DATA | CMD_HOLE_PT,
+		  uffd_data_handler, uffd_pt_handler, 2),
+
+	/*
+	 * Try accesses when the data and PT memory regions are both
+	 * tracked for dirty logging.
+	 */
+	TEST_DIRTY_LOG(guest_read64, with_af, guest_check_no_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_read64, no_af, guest_check_no_write_in_dirty_log,
+		       guest_check_no_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_ld_preidx, with_af,
+		       guest_check_no_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_at, no_af, guest_check_no_write_in_dirty_log,
+		       guest_check_no_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_exec, with_af, guest_check_no_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_write64, with_af, guest_check_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_cas, with_af, guest_check_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_dc_zva, with_af, guest_check_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+	TEST_DIRTY_LOG(guest_st_preidx, with_af, guest_check_write_in_dirty_log,
+		       guest_check_s1ptw_wr_in_dirty_log),
+
+	/*
+	 * Access when the data and PT memory regions are both marked for
+	 * dirty logging and UFFD at the same time. The expected result is
+	 * that writes should mark the dirty log and trigger a userfaultfd
+	 * write fault.  Reads/execs should result in a read userfaultfd
+	 * fault, and nothing in the dirty log.  Any S1PTW should result in
+	 * a write in the dirty log and a userfaultfd write.
+	 */
+	TEST_UFFD_AND_DIRTY_LOG(guest_read64, with_af,
+				uffd_data_handler, 2,
+				guest_check_no_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_read64, no_af,
+				uffd_data_handler, 2,
+				guest_check_no_write_in_dirty_log,
+				guest_check_no_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_ld_preidx, with_af,
+				uffd_data_handler,
+				2, guest_check_no_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_at, with_af, uffd_no_handler, 1,
+				guest_check_no_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_exec, with_af,
+				uffd_data_handler, 2,
+				guest_check_no_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_write64, with_af,
+				uffd_data_handler,
+				2, guest_check_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_cas, with_af,
+				uffd_data_handler, 2,
+				guest_check_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_dc_zva, with_af,
+				uffd_data_handler,
+				2, guest_check_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	TEST_UFFD_AND_DIRTY_LOG(guest_st_preidx, with_af,
+				uffd_data_handler, 2,
+				guest_check_write_in_dirty_log,
+				guest_check_s1ptw_wr_in_dirty_log),
+	/*
+	 * Access when both the PT and data regions are marked read-only
+	 * (with KVM_MEM_READONLY). Writes with a syndrome result in an
+	 * MMIO exit, writes with no syndrome (e.g., CAS) result in a
+	 * failed vcpu run, and reads/execs with and without syndroms do
+	 * not fault.
+	 */
+	TEST_RO_MEMSLOT(guest_read64, 0, 0),
+	TEST_RO_MEMSLOT(guest_ld_preidx, 0, 0),
+	TEST_RO_MEMSLOT(guest_at, 0, 0),
+	TEST_RO_MEMSLOT(guest_exec, 0, 0),
+	TEST_RO_MEMSLOT(guest_write64, mmio_on_test_gpa_handler, 1),
+	TEST_RO_MEMSLOT_NO_SYNDROME(guest_dc_zva),
+	TEST_RO_MEMSLOT_NO_SYNDROME(guest_cas),
+	TEST_RO_MEMSLOT_NO_SYNDROME(guest_st_preidx),
+
+	/*
+	 * The PT and data regions are both read-only and marked
+	 * for dirty logging at the same time. The expected result is that
+	 * for writes there should be no write in the dirty log. The
+	 * readonly handling is the same as if the memslot was not marked
+	 * for dirty logging: writes with a syndrome result in an MMIO
+	 * exit, and writes with no syndrome result in a failed vcpu run.
+	 */
+	TEST_RO_MEMSLOT_AND_DIRTY_LOG(guest_read64, 0, 0,
+				      guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_AND_DIRTY_LOG(guest_ld_preidx, 0, 0,
+				      guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_AND_DIRTY_LOG(guest_at, 0, 0,
+				      guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_AND_DIRTY_LOG(guest_exec, 0, 0,
+				      guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_AND_DIRTY_LOG(guest_write64, mmio_on_test_gpa_handler,
+				      1, guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_DIRTY_LOG(guest_dc_zva,
+						  guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_DIRTY_LOG(guest_cas,
+						  guest_check_no_write_in_dirty_log),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_DIRTY_LOG(guest_st_preidx,
+						  guest_check_no_write_in_dirty_log),
+
+	/*
+	 * The PT and data regions are both read-only and punched with
+	 * holes tracked with userfaultfd.  The expected result is the
+	 * union of both userfaultfd and read-only behaviors. For example,
+	 * write accesses result in a userfaultfd write fault and an MMIO
+	 * exit.  Writes with no syndrome result in a failed vcpu run and
+	 * no userfaultfd write fault. Reads result in userfaultfd getting
+	 * triggered.
+	 */
+	TEST_RO_MEMSLOT_AND_UFFD(guest_read64, 0, 0, uffd_data_handler, 2),
+	TEST_RO_MEMSLOT_AND_UFFD(guest_ld_preidx, 0, 0, uffd_data_handler, 2),
+	TEST_RO_MEMSLOT_AND_UFFD(guest_at, 0, 0, uffd_no_handler, 1),
+	TEST_RO_MEMSLOT_AND_UFFD(guest_exec, 0, 0, uffd_data_handler, 2),
+	TEST_RO_MEMSLOT_AND_UFFD(guest_write64, mmio_on_test_gpa_handler, 1,
+				 uffd_data_handler, 2),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_UFFD(guest_cas, uffd_data_handler, 2),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_UFFD(guest_dc_zva, uffd_no_handler, 1),
+	TEST_RO_MEMSLOT_NO_SYNDROME_AND_UFFD(guest_st_preidx, uffd_no_handler, 1),
+
+	{ 0 }
+};
+
+static void for_each_test_and_guest_mode(enum vm_mem_backing_src_type src_type)
+{
+	struct test_desc *t;
+
+	for (t = &tests[0]; t->name; t++) {
+		if (t->skip)
+			continue;
+
+		struct test_params p = {
+			.src_type = src_type,
+			.test_desc = t,
+		};
+
+		for_each_guest_mode(run_test, &p);
+	}
+}
+
+int main(int argc, char *argv[])
+{
+	enum vm_mem_backing_src_type src_type;
+	int opt;
+
+	src_type = DEFAULT_VM_MEM_SRC;
+
+	while ((opt = getopt(argc, argv, "hm:s:")) != -1) {
+		switch (opt) {
+		case 'm':
+			guest_modes_cmdline(optarg);
+			break;
+		case 's':
+			src_type = parse_backing_src_type(optarg);
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			exit(0);
+		}
+	}
+
+	for_each_test_and_guest_mode(src_type);
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/psci_test.c b/tools/testing/selftests/kvm/aarch64/psci_test.c
new file mode 100644
index 0000000000..9b004905d1
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/psci_test.c
@@ -0,0 +1,198 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * psci_test - Tests relating to KVM's PSCI implementation.
+ *
+ * Copyright (c) 2021 Google LLC.
+ *
+ * This test includes:
+ *  - A regression test for a race between KVM servicing the PSCI CPU_ON call
+ *    and userspace reading the targeted vCPU's registers.
+ *  - A test for KVM's handling of PSCI SYSTEM_SUSPEND and the associated
+ *    KVM_SYSTEM_EVENT_SUSPEND UAPI.
+ */
+
+#define _GNU_SOURCE
+
+#include <linux/psci.h>
+
+#include "kvm_util.h"
+#include "processor.h"
+#include "test_util.h"
+
+#define CPU_ON_ENTRY_ADDR 0xfeedf00dul
+#define CPU_ON_CONTEXT_ID 0xdeadc0deul
+
+static uint64_t psci_cpu_on(uint64_t target_cpu, uint64_t entry_addr,
+			    uint64_t context_id)
+{
+	struct arm_smccc_res res;
+
+	smccc_hvc(PSCI_0_2_FN64_CPU_ON, target_cpu, entry_addr, context_id,
+		  0, 0, 0, 0, &res);
+
+	return res.a0;
+}
+
+static uint64_t psci_affinity_info(uint64_t target_affinity,
+				   uint64_t lowest_affinity_level)
+{
+	struct arm_smccc_res res;
+
+	smccc_hvc(PSCI_0_2_FN64_AFFINITY_INFO, target_affinity, lowest_affinity_level,
+		  0, 0, 0, 0, 0, &res);
+
+	return res.a0;
+}
+
+static uint64_t psci_system_suspend(uint64_t entry_addr, uint64_t context_id)
+{
+	struct arm_smccc_res res;
+
+	smccc_hvc(PSCI_1_0_FN64_SYSTEM_SUSPEND, entry_addr, context_id,
+		  0, 0, 0, 0, 0, &res);
+
+	return res.a0;
+}
+
+static uint64_t psci_features(uint32_t func_id)
+{
+	struct arm_smccc_res res;
+
+	smccc_hvc(PSCI_1_0_FN_PSCI_FEATURES, func_id, 0, 0, 0, 0, 0, 0, &res);
+
+	return res.a0;
+}
+
+static void vcpu_power_off(struct kvm_vcpu *vcpu)
+{
+	struct kvm_mp_state mp_state = {
+		.mp_state = KVM_MP_STATE_STOPPED,
+	};
+
+	vcpu_mp_state_set(vcpu, &mp_state);
+}
+
+static struct kvm_vm *setup_vm(void *guest_code, struct kvm_vcpu **source,
+			       struct kvm_vcpu **target)
+{
+	struct kvm_vcpu_init init;
+	struct kvm_vm *vm;
+
+	vm = vm_create(2);
+
+	vm_ioctl(vm, KVM_ARM_PREFERRED_TARGET, &init);
+	init.features[0] |= (1 << KVM_ARM_VCPU_PSCI_0_2);
+
+	*source = aarch64_vcpu_add(vm, 0, &init, guest_code);
+	*target = aarch64_vcpu_add(vm, 1, &init, guest_code);
+
+	return vm;
+}
+
+static void enter_guest(struct kvm_vcpu *vcpu)
+{
+	struct ucall uc;
+
+	vcpu_run(vcpu);
+	if (get_ucall(vcpu, &uc) == UCALL_ABORT)
+		REPORT_GUEST_ASSERT(uc);
+}
+
+static void assert_vcpu_reset(struct kvm_vcpu *vcpu)
+{
+	uint64_t obs_pc, obs_x0;
+
+	vcpu_get_reg(vcpu, ARM64_CORE_REG(regs.pc), &obs_pc);
+	vcpu_get_reg(vcpu, ARM64_CORE_REG(regs.regs[0]), &obs_x0);
+
+	TEST_ASSERT(obs_pc == CPU_ON_ENTRY_ADDR,
+		    "unexpected target cpu pc: %lx (expected: %lx)",
+		    obs_pc, CPU_ON_ENTRY_ADDR);
+	TEST_ASSERT(obs_x0 == CPU_ON_CONTEXT_ID,
+		    "unexpected target context id: %lx (expected: %lx)",
+		    obs_x0, CPU_ON_CONTEXT_ID);
+}
+
+static void guest_test_cpu_on(uint64_t target_cpu)
+{
+	uint64_t target_state;
+
+	GUEST_ASSERT(!psci_cpu_on(target_cpu, CPU_ON_ENTRY_ADDR, CPU_ON_CONTEXT_ID));
+
+	do {
+		target_state = psci_affinity_info(target_cpu, 0);
+
+		GUEST_ASSERT((target_state == PSCI_0_2_AFFINITY_LEVEL_ON) ||
+			     (target_state == PSCI_0_2_AFFINITY_LEVEL_OFF));
+	} while (target_state != PSCI_0_2_AFFINITY_LEVEL_ON);
+
+	GUEST_DONE();
+}
+
+static void host_test_cpu_on(void)
+{
+	struct kvm_vcpu *source, *target;
+	uint64_t target_mpidr;
+	struct kvm_vm *vm;
+	struct ucall uc;
+
+	vm = setup_vm(guest_test_cpu_on, &source, &target);
+
+	/*
+	 * make sure the target is already off when executing the test.
+	 */
+	vcpu_power_off(target);
+
+	vcpu_get_reg(target, KVM_ARM64_SYS_REG(SYS_MPIDR_EL1), &target_mpidr);
+	vcpu_args_set(source, 1, target_mpidr & MPIDR_HWID_BITMASK);
+	enter_guest(source);
+
+	if (get_ucall(source, &uc) != UCALL_DONE)
+		TEST_FAIL("Unhandled ucall: %lu", uc.cmd);
+
+	assert_vcpu_reset(target);
+	kvm_vm_free(vm);
+}
+
+static void guest_test_system_suspend(void)
+{
+	uint64_t ret;
+
+	/* assert that SYSTEM_SUSPEND is discoverable */
+	GUEST_ASSERT(!psci_features(PSCI_1_0_FN_SYSTEM_SUSPEND));
+	GUEST_ASSERT(!psci_features(PSCI_1_0_FN64_SYSTEM_SUSPEND));
+
+	ret = psci_system_suspend(CPU_ON_ENTRY_ADDR, CPU_ON_CONTEXT_ID);
+	GUEST_SYNC(ret);
+}
+
+static void host_test_system_suspend(void)
+{
+	struct kvm_vcpu *source, *target;
+	struct kvm_run *run;
+	struct kvm_vm *vm;
+
+	vm = setup_vm(guest_test_system_suspend, &source, &target);
+	vm_enable_cap(vm, KVM_CAP_ARM_SYSTEM_SUSPEND, 0);
+
+	vcpu_power_off(target);
+	run = source->run;
+
+	enter_guest(source);
+
+	TEST_ASSERT_KVM_EXIT_REASON(source, KVM_EXIT_SYSTEM_EVENT);
+	TEST_ASSERT(run->system_event.type == KVM_SYSTEM_EVENT_SUSPEND,
+		    "Unhandled system event: %u (expected: %u)",
+		    run->system_event.type, KVM_SYSTEM_EVENT_SUSPEND);
+
+	kvm_vm_free(vm);
+}
+
+int main(void)
+{
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_ARM_SYSTEM_SUSPEND));
+
+	host_test_cpu_on();
+	host_test_system_suspend();
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/smccc_filter.c b/tools/testing/selftests/kvm/aarch64/smccc_filter.c
new file mode 100644
index 0000000000..f4ceae9c89
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/smccc_filter.c
@@ -0,0 +1,268 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * smccc_filter - Tests for the SMCCC filter UAPI.
+ *
+ * Copyright (c) 2023 Google LLC
+ *
+ * This test includes:
+ *  - Tests that the UAPI constraints are upheld by KVM. For example, userspace
+ *    is prevented from filtering the architecture range of SMCCC calls.
+ *  - Test that the filter actions (DENIED, FWD_TO_USER) work as intended.
+ */
+
+#include <linux/arm-smccc.h>
+#include <linux/psci.h>
+#include <stdint.h>
+
+#include "processor.h"
+#include "test_util.h"
+
+enum smccc_conduit {
+	HVC_INSN,
+	SMC_INSN,
+};
+
+#define for_each_conduit(conduit)					\
+	for (conduit = HVC_INSN; conduit <= SMC_INSN; conduit++)
+
+static void guest_main(uint32_t func_id, enum smccc_conduit conduit)
+{
+	struct arm_smccc_res res;
+
+	if (conduit == SMC_INSN)
+		smccc_smc(func_id, 0, 0, 0, 0, 0, 0, 0, &res);
+	else
+		smccc_hvc(func_id, 0, 0, 0, 0, 0, 0, 0, &res);
+
+	GUEST_SYNC(res.a0);
+}
+
+static int __set_smccc_filter(struct kvm_vm *vm, uint32_t start, uint32_t nr_functions,
+			      enum kvm_smccc_filter_action action)
+{
+	struct kvm_smccc_filter filter = {
+		.base		= start,
+		.nr_functions	= nr_functions,
+		.action		= action,
+	};
+
+	return __kvm_device_attr_set(vm->fd, KVM_ARM_VM_SMCCC_CTRL,
+				     KVM_ARM_VM_SMCCC_FILTER, &filter);
+}
+
+static void set_smccc_filter(struct kvm_vm *vm, uint32_t start, uint32_t nr_functions,
+			     enum kvm_smccc_filter_action action)
+{
+	int ret = __set_smccc_filter(vm, start, nr_functions, action);
+
+	TEST_ASSERT(!ret, "failed to configure SMCCC filter: %d", ret);
+}
+
+static struct kvm_vm *setup_vm(struct kvm_vcpu **vcpu)
+{
+	struct kvm_vcpu_init init;
+	struct kvm_vm *vm;
+
+	vm = vm_create(1);
+	vm_ioctl(vm, KVM_ARM_PREFERRED_TARGET, &init);
+
+	/*
+	 * Enable in-kernel emulation of PSCI to ensure that calls are denied
+	 * due to the SMCCC filter, not because of KVM.
+	 */
+	init.features[0] |= (1 << KVM_ARM_VCPU_PSCI_0_2);
+
+	*vcpu = aarch64_vcpu_add(vm, 0, &init, guest_main);
+	return vm;
+}
+
+static void test_pad_must_be_zero(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	struct kvm_smccc_filter filter = {
+		.base		= PSCI_0_2_FN_PSCI_VERSION,
+		.nr_functions	= 1,
+		.action		= KVM_SMCCC_FILTER_DENY,
+		.pad		= { -1 },
+	};
+	int r;
+
+	r = __kvm_device_attr_set(vm->fd, KVM_ARM_VM_SMCCC_CTRL,
+				  KVM_ARM_VM_SMCCC_FILTER, &filter);
+	TEST_ASSERT(r < 0 && errno == EINVAL,
+		    "Setting filter with nonzero padding should return EINVAL");
+}
+
+/* Ensure that userspace cannot filter the Arm Architecture SMCCC range */
+static void test_filter_reserved_range(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	uint32_t smc64_fn;
+	int r;
+
+	r = __set_smccc_filter(vm, ARM_SMCCC_ARCH_WORKAROUND_1,
+			       1, KVM_SMCCC_FILTER_DENY);
+	TEST_ASSERT(r < 0 && errno == EEXIST,
+		    "Attempt to filter reserved range should return EEXIST");
+
+	smc64_fn = ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, ARM_SMCCC_SMC_64,
+				      0, 0);
+
+	r = __set_smccc_filter(vm, smc64_fn, 1, KVM_SMCCC_FILTER_DENY);
+	TEST_ASSERT(r < 0 && errno == EEXIST,
+		    "Attempt to filter reserved range should return EEXIST");
+
+	kvm_vm_free(vm);
+}
+
+static void test_invalid_nr_functions(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	int r;
+
+	r = __set_smccc_filter(vm, PSCI_0_2_FN64_CPU_ON, 0, KVM_SMCCC_FILTER_DENY);
+	TEST_ASSERT(r < 0 && errno == EINVAL,
+		    "Attempt to filter 0 functions should return EINVAL");
+
+	kvm_vm_free(vm);
+}
+
+static void test_overflow_nr_functions(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	int r;
+
+	r = __set_smccc_filter(vm, ~0, ~0, KVM_SMCCC_FILTER_DENY);
+	TEST_ASSERT(r < 0 && errno == EINVAL,
+		    "Attempt to overflow filter range should return EINVAL");
+
+	kvm_vm_free(vm);
+}
+
+static void test_reserved_action(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	int r;
+
+	r = __set_smccc_filter(vm, PSCI_0_2_FN64_CPU_ON, 1, -1);
+	TEST_ASSERT(r < 0 && errno == EINVAL,
+		    "Attempt to use reserved filter action should return EINVAL");
+
+	kvm_vm_free(vm);
+}
+
+
+/* Test that overlapping configurations of the SMCCC filter are rejected */
+static void test_filter_overlap(void)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm = setup_vm(&vcpu);
+	int r;
+
+	set_smccc_filter(vm, PSCI_0_2_FN64_CPU_ON, 1, KVM_SMCCC_FILTER_DENY);
+
+	r = __set_smccc_filter(vm, PSCI_0_2_FN64_CPU_ON, 1, KVM_SMCCC_FILTER_DENY);
+	TEST_ASSERT(r < 0 && errno == EEXIST,
+		    "Attempt to filter already configured range should return EEXIST");
+
+	kvm_vm_free(vm);
+}
+
+static void expect_call_denied(struct kvm_vcpu *vcpu)
+{
+	struct ucall uc;
+
+	if (get_ucall(vcpu, &uc) != UCALL_SYNC)
+		TEST_FAIL("Unexpected ucall: %lu\n", uc.cmd);
+
+	TEST_ASSERT(uc.args[1] == SMCCC_RET_NOT_SUPPORTED,
+		    "Unexpected SMCCC return code: %lu", uc.args[1]);
+}
+
+/* Denied SMCCC calls have a return code of SMCCC_RET_NOT_SUPPORTED */
+static void test_filter_denied(void)
+{
+	enum smccc_conduit conduit;
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+
+	for_each_conduit(conduit) {
+		vm = setup_vm(&vcpu);
+
+		set_smccc_filter(vm, PSCI_0_2_FN_PSCI_VERSION, 1, KVM_SMCCC_FILTER_DENY);
+		vcpu_args_set(vcpu, 2, PSCI_0_2_FN_PSCI_VERSION, conduit);
+
+		vcpu_run(vcpu);
+		expect_call_denied(vcpu);
+
+		kvm_vm_free(vm);
+	}
+}
+
+static void expect_call_fwd_to_user(struct kvm_vcpu *vcpu, uint32_t func_id,
+				    enum smccc_conduit conduit)
+{
+	struct kvm_run *run = vcpu->run;
+
+	TEST_ASSERT(run->exit_reason == KVM_EXIT_HYPERCALL,
+		    "Unexpected exit reason: %u", run->exit_reason);
+	TEST_ASSERT(run->hypercall.nr == func_id,
+		    "Unexpected SMCCC function: %llu", run->hypercall.nr);
+
+	if (conduit == SMC_INSN)
+		TEST_ASSERT(run->hypercall.flags & KVM_HYPERCALL_EXIT_SMC,
+			    "KVM_HYPERCALL_EXIT_SMC is not set");
+	else
+		TEST_ASSERT(!(run->hypercall.flags & KVM_HYPERCALL_EXIT_SMC),
+			    "KVM_HYPERCALL_EXIT_SMC is set");
+}
+
+/* SMCCC calls forwarded to userspace cause KVM_EXIT_HYPERCALL exits */
+static void test_filter_fwd_to_user(void)
+{
+	enum smccc_conduit conduit;
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+
+	for_each_conduit(conduit) {
+		vm = setup_vm(&vcpu);
+
+		set_smccc_filter(vm, PSCI_0_2_FN_PSCI_VERSION, 1, KVM_SMCCC_FILTER_FWD_TO_USER);
+		vcpu_args_set(vcpu, 2, PSCI_0_2_FN_PSCI_VERSION, conduit);
+
+		vcpu_run(vcpu);
+		expect_call_fwd_to_user(vcpu, PSCI_0_2_FN_PSCI_VERSION, conduit);
+
+		kvm_vm_free(vm);
+	}
+}
+
+static bool kvm_supports_smccc_filter(void)
+{
+	struct kvm_vm *vm = vm_create_barebones();
+	int r;
+
+	r = __kvm_has_device_attr(vm->fd, KVM_ARM_VM_SMCCC_CTRL, KVM_ARM_VM_SMCCC_FILTER);
+
+	kvm_vm_free(vm);
+	return !r;
+}
+
+int main(void)
+{
+	TEST_REQUIRE(kvm_supports_smccc_filter());
+
+	test_pad_must_be_zero();
+	test_invalid_nr_functions();
+	test_overflow_nr_functions();
+	test_reserved_action();
+	test_filter_reserved_range();
+	test_filter_overlap();
+	test_filter_denied();
+	test_filter_fwd_to_user();
+}
diff --git a/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c b/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c
new file mode 100644
index 0000000000..80b74c6f15
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c
@@ -0,0 +1,121 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * vcpu_width_config - Test KVM_ARM_VCPU_INIT() with KVM_ARM_VCPU_EL1_32BIT.
+ *
+ * Copyright (c) 2022 Google LLC.
+ *
+ * This is a test that ensures that non-mixed-width vCPUs (all 64bit vCPUs
+ * or all 32bit vcPUs) can be configured and mixed-width vCPUs cannot be
+ * configured.
+ */
+
+#include "kvm_util.h"
+#include "processor.h"
+#include "test_util.h"
+
+
+/*
+ * Add a vCPU, run KVM_ARM_VCPU_INIT with @init0, and then
+ * add another vCPU, and run KVM_ARM_VCPU_INIT with @init1.
+ */
+static int add_init_2vcpus(struct kvm_vcpu_init *init0,
+			   struct kvm_vcpu_init *init1)
+{
+	struct kvm_vcpu *vcpu0, *vcpu1;
+	struct kvm_vm *vm;
+	int ret;
+
+	vm = vm_create_barebones();
+
+	vcpu0 = __vm_vcpu_add(vm, 0);
+	ret = __vcpu_ioctl(vcpu0, KVM_ARM_VCPU_INIT, init0);
+	if (ret)
+		goto free_exit;
+
+	vcpu1 = __vm_vcpu_add(vm, 1);
+	ret = __vcpu_ioctl(vcpu1, KVM_ARM_VCPU_INIT, init1);
+
+free_exit:
+	kvm_vm_free(vm);
+	return ret;
+}
+
+/*
+ * Add two vCPUs, then run KVM_ARM_VCPU_INIT for one vCPU with @init0,
+ * and run KVM_ARM_VCPU_INIT for another vCPU with @init1.
+ */
+static int add_2vcpus_init_2vcpus(struct kvm_vcpu_init *init0,
+				  struct kvm_vcpu_init *init1)
+{
+	struct kvm_vcpu *vcpu0, *vcpu1;
+	struct kvm_vm *vm;
+	int ret;
+
+	vm = vm_create_barebones();
+
+	vcpu0 = __vm_vcpu_add(vm, 0);
+	vcpu1 = __vm_vcpu_add(vm, 1);
+
+	ret = __vcpu_ioctl(vcpu0, KVM_ARM_VCPU_INIT, init0);
+	if (ret)
+		goto free_exit;
+
+	ret = __vcpu_ioctl(vcpu1, KVM_ARM_VCPU_INIT, init1);
+
+free_exit:
+	kvm_vm_free(vm);
+	return ret;
+}
+
+/*
+ * Tests that two 64bit vCPUs can be configured, two 32bit vCPUs can be
+ * configured, and two mixed-width vCPUs cannot be configured.
+ * Each of those three cases, configure vCPUs in two different orders.
+ * The one is running KVM_CREATE_VCPU for 2 vCPUs, and then running
+ * KVM_ARM_VCPU_INIT for them.
+ * The other is running KVM_CREATE_VCPU and KVM_ARM_VCPU_INIT for a vCPU,
+ * and then run those commands for another vCPU.
+ */
+int main(void)
+{
+	struct kvm_vcpu_init init0, init1;
+	struct kvm_vm *vm;
+	int ret;
+
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_ARM_EL1_32BIT));
+
+	/* Get the preferred target type and copy that to init1 for later use */
+	vm = vm_create_barebones();
+	vm_ioctl(vm, KVM_ARM_PREFERRED_TARGET, &init0);
+	kvm_vm_free(vm);
+	init1 = init0;
+
+	/* Test with 64bit vCPUs */
+	ret = add_init_2vcpus(&init0, &init0);
+	TEST_ASSERT(ret == 0,
+		    "Configuring 64bit EL1 vCPUs failed unexpectedly");
+	ret = add_2vcpus_init_2vcpus(&init0, &init0);
+	TEST_ASSERT(ret == 0,
+		    "Configuring 64bit EL1 vCPUs failed unexpectedly");
+
+	/* Test with 32bit vCPUs */
+	init0.features[0] = (1 << KVM_ARM_VCPU_EL1_32BIT);
+	ret = add_init_2vcpus(&init0, &init0);
+	TEST_ASSERT(ret == 0,
+		    "Configuring 32bit EL1 vCPUs failed unexpectedly");
+	ret = add_2vcpus_init_2vcpus(&init0, &init0);
+	TEST_ASSERT(ret == 0,
+		    "Configuring 32bit EL1 vCPUs failed unexpectedly");
+
+	/* Test with mixed-width vCPUs  */
+	init0.features[0] = 0;
+	init1.features[0] = (1 << KVM_ARM_VCPU_EL1_32BIT);
+	ret = add_init_2vcpus(&init0, &init1);
+	TEST_ASSERT(ret != 0,
+		    "Configuring mixed-width vCPUs worked unexpectedly");
+	ret = add_2vcpus_init_2vcpus(&init0, &init1);
+	TEST_ASSERT(ret != 0,
+		    "Configuring mixed-width vCPUs worked unexpectedly");
+
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/vgic_init.c b/tools/testing/selftests/kvm/aarch64/vgic_init.c
new file mode 100644
index 0000000000..eef816b809
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/vgic_init.c
@@ -0,0 +1,716 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * vgic init sequence tests
+ *
+ * Copyright (C) 2020, Red Hat, Inc.
+ */
+#define _GNU_SOURCE
+#include <linux/kernel.h>
+#include <sys/syscall.h>
+#include <asm/kvm.h>
+#include <asm/kvm_para.h>
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "processor.h"
+#include "vgic.h"
+
+#define NR_VCPUS		4
+
+#define REG_OFFSET(vcpu, offset) (((uint64_t)vcpu << 32) | offset)
+
+#define GICR_TYPER 0x8
+
+#define VGIC_DEV_IS_V2(_d) ((_d) == KVM_DEV_TYPE_ARM_VGIC_V2)
+#define VGIC_DEV_IS_V3(_d) ((_d) == KVM_DEV_TYPE_ARM_VGIC_V3)
+
+struct vm_gic {
+	struct kvm_vm *vm;
+	int gic_fd;
+	uint32_t gic_dev_type;
+};
+
+static uint64_t max_phys_size;
+
+/*
+ * Helpers to access a redistributor register and verify the ioctl() failed or
+ * succeeded as expected, and provided the correct value on success.
+ */
+static void v3_redist_reg_get_errno(int gicv3_fd, int vcpu, int offset,
+				    int want, const char *msg)
+{
+	uint32_t ignored_val;
+	int ret = __kvm_device_attr_get(gicv3_fd, KVM_DEV_ARM_VGIC_GRP_REDIST_REGS,
+					REG_OFFSET(vcpu, offset), &ignored_val);
+
+	TEST_ASSERT(ret && errno == want, "%s; want errno = %d", msg, want);
+}
+
+static void v3_redist_reg_get(int gicv3_fd, int vcpu, int offset, uint32_t want,
+			      const char *msg)
+{
+	uint32_t val;
+
+	kvm_device_attr_get(gicv3_fd, KVM_DEV_ARM_VGIC_GRP_REDIST_REGS,
+			    REG_OFFSET(vcpu, offset), &val);
+	TEST_ASSERT(val == want, "%s; want '0x%x', got '0x%x'", msg, want, val);
+}
+
+/* dummy guest code */
+static void guest_code(void)
+{
+	GUEST_SYNC(0);
+	GUEST_SYNC(1);
+	GUEST_SYNC(2);
+	GUEST_DONE();
+}
+
+/* we don't want to assert on run execution, hence that helper */
+static int run_vcpu(struct kvm_vcpu *vcpu)
+{
+	return __vcpu_run(vcpu) ? -errno : 0;
+}
+
+static struct vm_gic vm_gic_create_with_vcpus(uint32_t gic_dev_type,
+					      uint32_t nr_vcpus,
+					      struct kvm_vcpu *vcpus[])
+{
+	struct vm_gic v;
+
+	v.gic_dev_type = gic_dev_type;
+	v.vm = vm_create_with_vcpus(nr_vcpus, guest_code, vcpus);
+	v.gic_fd = kvm_create_device(v.vm, gic_dev_type);
+
+	return v;
+}
+
+static void vm_gic_destroy(struct vm_gic *v)
+{
+	close(v->gic_fd);
+	kvm_vm_free(v->vm);
+}
+
+struct vgic_region_attr {
+	uint64_t attr;
+	uint64_t size;
+	uint64_t alignment;
+};
+
+struct vgic_region_attr gic_v3_dist_region = {
+	.attr = KVM_VGIC_V3_ADDR_TYPE_DIST,
+	.size = 0x10000,
+	.alignment = 0x10000,
+};
+
+struct vgic_region_attr gic_v3_redist_region = {
+	.attr = KVM_VGIC_V3_ADDR_TYPE_REDIST,
+	.size = NR_VCPUS * 0x20000,
+	.alignment = 0x10000,
+};
+
+struct vgic_region_attr gic_v2_dist_region = {
+	.attr = KVM_VGIC_V2_ADDR_TYPE_DIST,
+	.size = 0x1000,
+	.alignment = 0x1000,
+};
+
+struct vgic_region_attr gic_v2_cpu_region = {
+	.attr = KVM_VGIC_V2_ADDR_TYPE_CPU,
+	.size = 0x2000,
+	.alignment = 0x1000,
+};
+
+/**
+ * Helper routine that performs KVM device tests in general. Eventually the
+ * ARM_VGIC (GICv2 or GICv3) device gets created with an overlapping
+ * DIST/REDIST (or DIST/CPUIF for GICv2). Assumption is 4 vcpus are going to be
+ * used hence the overlap. In the case of GICv3, A RDIST region is set at @0x0
+ * and a DIST region is set @0x70000. The GICv2 case sets a CPUIF @0x0 and a
+ * DIST region @0x1000.
+ */
+static void subtest_dist_rdist(struct vm_gic *v)
+{
+	int ret;
+	uint64_t addr;
+	struct vgic_region_attr rdist; /* CPU interface in GICv2*/
+	struct vgic_region_attr dist;
+
+	rdist = VGIC_DEV_IS_V3(v->gic_dev_type) ? gic_v3_redist_region
+						: gic_v2_cpu_region;
+	dist = VGIC_DEV_IS_V3(v->gic_dev_type) ? gic_v3_dist_region
+						: gic_v2_dist_region;
+
+	/* Check existing group/attributes */
+	kvm_has_device_attr(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR, dist.attr);
+
+	kvm_has_device_attr(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR, rdist.attr);
+
+	/* check non existing attribute */
+	ret = __kvm_has_device_attr(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR, -1);
+	TEST_ASSERT(ret && errno == ENXIO, "attribute not supported");
+
+	/* misaligned DIST and REDIST address settings */
+	addr = dist.alignment / 0x10;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    dist.attr, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "GIC dist base not aligned");
+
+	addr = rdist.alignment / 0x10;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    rdist.attr, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "GIC redist/cpu base not aligned");
+
+	/* out of range address */
+	addr = max_phys_size;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    dist.attr, &addr);
+	TEST_ASSERT(ret && errno == E2BIG, "dist address beyond IPA limit");
+
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    rdist.attr, &addr);
+	TEST_ASSERT(ret && errno == E2BIG, "redist address beyond IPA limit");
+
+	/* Space for half a rdist (a rdist is: 2 * rdist.alignment). */
+	addr = max_phys_size - dist.alignment;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    rdist.attr, &addr);
+	TEST_ASSERT(ret && errno == E2BIG,
+			"half of the redist is beyond IPA limit");
+
+	/* set REDIST base address @0x0*/
+	addr = 0x00000;
+	kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    rdist.attr, &addr);
+
+	/* Attempt to create a second legacy redistributor region */
+	addr = 0xE0000;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    rdist.attr, &addr);
+	TEST_ASSERT(ret && errno == EEXIST, "GIC redist base set again");
+
+	ret = __kvm_has_device_attr(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				     KVM_VGIC_V3_ADDR_TYPE_REDIST);
+	if (!ret) {
+		/* Attempt to mix legacy and new redistributor regions */
+		addr = REDIST_REGION_ATTR_ADDR(NR_VCPUS, 0x100000, 0, 0);
+		ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+					    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+		TEST_ASSERT(ret && errno == EINVAL,
+			    "attempt to mix GICv3 REDIST and REDIST_REGION");
+	}
+
+	/*
+	 * Set overlapping DIST / REDIST, cannot be detected here. Will be detected
+	 * on first vcpu run instead.
+	 */
+	addr = rdist.size - rdist.alignment;
+	kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    dist.attr, &addr);
+}
+
+/* Test the new REDIST region API */
+static void subtest_v3_redist_regions(struct vm_gic *v)
+{
+	uint64_t addr, expected_addr;
+	int ret;
+
+	ret = __kvm_has_device_attr(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST);
+	TEST_ASSERT(!ret, "Multiple redist regions advertised");
+
+	addr = REDIST_REGION_ATTR_ADDR(NR_VCPUS, 0x100000, 2, 0);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "redist region attr value with flags != 0");
+
+	addr = REDIST_REGION_ATTR_ADDR(0, 0x100000, 0, 0);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "redist region attr value with count== 0");
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 1);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL,
+		    "attempt to register the first rdist region with index != 0");
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x201000, 0, 1);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "rdist region with misaligned address");
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 0);
+	kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 1);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "register an rdist region with already used index");
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x210000, 0, 2);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL,
+		    "register an rdist region overlapping with another one");
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x240000, 0, 2);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "register redist region with index not +1");
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x240000, 0, 1);
+	kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	addr = REDIST_REGION_ATTR_ADDR(1, max_phys_size, 0, 2);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == E2BIG,
+		    "register redist region with base address beyond IPA range");
+
+	/* The last redist is above the pa range. */
+	addr = REDIST_REGION_ATTR_ADDR(2, max_phys_size - 0x30000, 0, 2);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == E2BIG,
+		    "register redist region with top address beyond IPA range");
+
+	addr = 0x260000;
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST, &addr);
+	TEST_ASSERT(ret && errno == EINVAL,
+		    "Mix KVM_VGIC_V3_ADDR_TYPE_REDIST and REDIST_REGION");
+
+	/*
+	 * Now there are 2 redist regions:
+	 * region 0 @ 0x200000 2 redists
+	 * region 1 @ 0x240000 1 redist
+	 * Attempt to read their characteristics
+	 */
+
+	addr = REDIST_REGION_ATTR_ADDR(0, 0, 0, 0);
+	expected_addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 0);
+	ret = __kvm_device_attr_get(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(!ret && addr == expected_addr, "read characteristics of region #0");
+
+	addr = REDIST_REGION_ATTR_ADDR(0, 0, 0, 1);
+	expected_addr = REDIST_REGION_ATTR_ADDR(1, 0x240000, 0, 1);
+	ret = __kvm_device_attr_get(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(!ret && addr == expected_addr, "read characteristics of region #1");
+
+	addr = REDIST_REGION_ATTR_ADDR(0, 0, 0, 2);
+	ret = __kvm_device_attr_get(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == ENOENT, "read characteristics of non existing region");
+
+	addr = 0x260000;
+	kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_DIST, &addr);
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x260000, 0, 2);
+	ret = __kvm_device_attr_set(v->gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "register redist region colliding with dist");
+}
+
+/*
+ * VGIC KVM device is created and initialized before the secondary CPUs
+ * get created
+ */
+static void test_vgic_then_vcpus(uint32_t gic_dev_type)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	struct vm_gic v;
+	int ret, i;
+
+	v = vm_gic_create_with_vcpus(gic_dev_type, 1, vcpus);
+
+	subtest_dist_rdist(&v);
+
+	/* Add the rest of the VCPUs */
+	for (i = 1; i < NR_VCPUS; ++i)
+		vcpus[i] = vm_vcpu_add(v.vm, i, guest_code);
+
+	ret = run_vcpu(vcpus[3]);
+	TEST_ASSERT(ret == -EINVAL, "dist/rdist overlap detected on 1st vcpu run");
+
+	vm_gic_destroy(&v);
+}
+
+/* All the VCPUs are created before the VGIC KVM device gets initialized */
+static void test_vcpus_then_vgic(uint32_t gic_dev_type)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	struct vm_gic v;
+	int ret;
+
+	v = vm_gic_create_with_vcpus(gic_dev_type, NR_VCPUS, vcpus);
+
+	subtest_dist_rdist(&v);
+
+	ret = run_vcpu(vcpus[3]);
+	TEST_ASSERT(ret == -EINVAL, "dist/rdist overlap detected on 1st vcpu run");
+
+	vm_gic_destroy(&v);
+}
+
+static void test_v3_new_redist_regions(void)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	void *dummy = NULL;
+	struct vm_gic v;
+	uint64_t addr;
+	int ret;
+
+	v = vm_gic_create_with_vcpus(KVM_DEV_TYPE_ARM_VGIC_V3, NR_VCPUS, vcpus);
+	subtest_v3_redist_regions(&v);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	ret = run_vcpu(vcpus[3]);
+	TEST_ASSERT(ret == -ENXIO, "running without sufficient number of rdists");
+	vm_gic_destroy(&v);
+
+	/* step2 */
+
+	v = vm_gic_create_with_vcpus(KVM_DEV_TYPE_ARM_VGIC_V3, NR_VCPUS, vcpus);
+	subtest_v3_redist_regions(&v);
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x280000, 0, 2);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	ret = run_vcpu(vcpus[3]);
+	TEST_ASSERT(ret == -EBUSY, "running without vgic explicit init");
+
+	vm_gic_destroy(&v);
+
+	/* step 3 */
+
+	v = vm_gic_create_with_vcpus(KVM_DEV_TYPE_ARM_VGIC_V3, NR_VCPUS, vcpus);
+	subtest_v3_redist_regions(&v);
+
+	ret = __kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, dummy);
+	TEST_ASSERT(ret && errno == EFAULT,
+		    "register a third region allowing to cover the 4 vcpus");
+
+	addr = REDIST_REGION_ATTR_ADDR(1, 0x280000, 0, 2);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	ret = run_vcpu(vcpus[3]);
+	TEST_ASSERT(!ret, "vcpu run");
+
+	vm_gic_destroy(&v);
+}
+
+static void test_v3_typer_accesses(void)
+{
+	struct vm_gic v;
+	uint64_t addr;
+	int ret, i;
+
+	v.vm = vm_create(NR_VCPUS);
+	(void)vm_vcpu_add(v.vm, 0, guest_code);
+
+	v.gic_fd = kvm_create_device(v.vm, KVM_DEV_TYPE_ARM_VGIC_V3);
+
+	(void)vm_vcpu_add(v.vm, 3, guest_code);
+
+	v3_redist_reg_get_errno(v.gic_fd, 1, GICR_TYPER, EINVAL,
+				"attempting to read GICR_TYPER of non created vcpu");
+
+	(void)vm_vcpu_add(v.vm, 1, guest_code);
+
+	v3_redist_reg_get_errno(v.gic_fd, 1, GICR_TYPER, EBUSY,
+				"read GICR_TYPER before GIC initialized");
+
+	(void)vm_vcpu_add(v.vm, 2, guest_code);
+
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	for (i = 0; i < NR_VCPUS ; i++) {
+		v3_redist_reg_get(v.gic_fd, i, GICR_TYPER, i * 0x100,
+				  "read GICR_TYPER before rdist region setting");
+	}
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 0);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	/* The 2 first rdists should be put there (vcpu 0 and 3) */
+	v3_redist_reg_get(v.gic_fd, 0, GICR_TYPER, 0x0, "read typer of rdist #0");
+	v3_redist_reg_get(v.gic_fd, 3, GICR_TYPER, 0x310, "read typer of rdist #1");
+
+	addr = REDIST_REGION_ATTR_ADDR(10, 0x100000, 0, 1);
+	ret = __kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+	TEST_ASSERT(ret && errno == EINVAL, "collision with previous rdist region");
+
+	v3_redist_reg_get(v.gic_fd, 1, GICR_TYPER, 0x100,
+			  "no redist region attached to vcpu #1 yet, last cannot be returned");
+	v3_redist_reg_get(v.gic_fd, 2, GICR_TYPER, 0x200,
+			  "no redist region attached to vcpu #2, last cannot be returned");
+
+	addr = REDIST_REGION_ATTR_ADDR(10, 0x20000, 0, 1);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	v3_redist_reg_get(v.gic_fd, 1, GICR_TYPER, 0x100, "read typer of rdist #1");
+	v3_redist_reg_get(v.gic_fd, 2, GICR_TYPER, 0x210,
+			  "read typer of rdist #1, last properly returned");
+
+	vm_gic_destroy(&v);
+}
+
+static struct vm_gic vm_gic_v3_create_with_vcpuids(int nr_vcpus,
+						   uint32_t vcpuids[])
+{
+	struct vm_gic v;
+	int i;
+
+	v.vm = vm_create(nr_vcpus);
+	for (i = 0; i < nr_vcpus; i++)
+		vm_vcpu_add(v.vm, vcpuids[i], guest_code);
+
+	v.gic_fd = kvm_create_device(v.vm, KVM_DEV_TYPE_ARM_VGIC_V3);
+
+	return v;
+}
+
+/**
+ * Test GICR_TYPER last bit with new redist regions
+ * rdist regions #1 and #2 are contiguous
+ * rdist region #0 @0x100000 2 rdist capacity
+ *     rdists: 0, 3 (Last)
+ * rdist region #1 @0x240000 2 rdist capacity
+ *     rdists:  5, 4 (Last)
+ * rdist region #2 @0x200000 2 rdist capacity
+ *     rdists: 1, 2
+ */
+static void test_v3_last_bit_redist_regions(void)
+{
+	uint32_t vcpuids[] = { 0, 3, 5, 4, 1, 2 };
+	struct vm_gic v;
+	uint64_t addr;
+
+	v = vm_gic_v3_create_with_vcpuids(ARRAY_SIZE(vcpuids), vcpuids);
+
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x100000, 0, 0);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x240000, 0, 1);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	addr = REDIST_REGION_ATTR_ADDR(2, 0x200000, 0, 2);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST_REGION, &addr);
+
+	v3_redist_reg_get(v.gic_fd, 0, GICR_TYPER, 0x000, "read typer of rdist #0");
+	v3_redist_reg_get(v.gic_fd, 1, GICR_TYPER, 0x100, "read typer of rdist #1");
+	v3_redist_reg_get(v.gic_fd, 2, GICR_TYPER, 0x200, "read typer of rdist #2");
+	v3_redist_reg_get(v.gic_fd, 3, GICR_TYPER, 0x310, "read typer of rdist #3");
+	v3_redist_reg_get(v.gic_fd, 5, GICR_TYPER, 0x500, "read typer of rdist #5");
+	v3_redist_reg_get(v.gic_fd, 4, GICR_TYPER, 0x410, "read typer of rdist #4");
+
+	vm_gic_destroy(&v);
+}
+
+/* Test last bit with legacy region */
+static void test_v3_last_bit_single_rdist(void)
+{
+	uint32_t vcpuids[] = { 0, 3, 5, 4, 1, 2 };
+	struct vm_gic v;
+	uint64_t addr;
+
+	v = vm_gic_v3_create_with_vcpuids(ARRAY_SIZE(vcpuids), vcpuids);
+
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	addr = 0x10000;
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST, &addr);
+
+	v3_redist_reg_get(v.gic_fd, 0, GICR_TYPER, 0x000, "read typer of rdist #0");
+	v3_redist_reg_get(v.gic_fd, 3, GICR_TYPER, 0x300, "read typer of rdist #1");
+	v3_redist_reg_get(v.gic_fd, 5, GICR_TYPER, 0x500, "read typer of rdist #2");
+	v3_redist_reg_get(v.gic_fd, 1, GICR_TYPER, 0x100, "read typer of rdist #3");
+	v3_redist_reg_get(v.gic_fd, 2, GICR_TYPER, 0x210, "read typer of rdist #3");
+
+	vm_gic_destroy(&v);
+}
+
+/* Uses the legacy REDIST region API. */
+static void test_v3_redist_ipa_range_check_at_vcpu_run(void)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	struct vm_gic v;
+	int ret, i;
+	uint64_t addr;
+
+	v = vm_gic_create_with_vcpus(KVM_DEV_TYPE_ARM_VGIC_V3, 1, vcpus);
+
+	/* Set space for 3 redists, we have 1 vcpu, so this succeeds. */
+	addr = max_phys_size - (3 * 2 * 0x10000);
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_REDIST, &addr);
+
+	addr = 0x00000;
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_V3_ADDR_TYPE_DIST, &addr);
+
+	/* Add the rest of the VCPUs */
+	for (i = 1; i < NR_VCPUS; ++i)
+		vcpus[i] = vm_vcpu_add(v.vm, i, guest_code);
+
+	kvm_device_attr_set(v.gic_fd, KVM_DEV_ARM_VGIC_GRP_CTRL,
+			    KVM_DEV_ARM_VGIC_CTRL_INIT, NULL);
+
+	/* Attempt to run a vcpu without enough redist space. */
+	ret = run_vcpu(vcpus[2]);
+	TEST_ASSERT(ret && errno == EINVAL,
+		"redist base+size above PA range detected on 1st vcpu run");
+
+	vm_gic_destroy(&v);
+}
+
+static void test_v3_its_region(void)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	struct vm_gic v;
+	uint64_t addr;
+	int its_fd, ret;
+
+	v = vm_gic_create_with_vcpus(KVM_DEV_TYPE_ARM_VGIC_V3, NR_VCPUS, vcpus);
+	its_fd = kvm_create_device(v.vm, KVM_DEV_TYPE_ARM_VGIC_ITS);
+
+	addr = 0x401000;
+	ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_ITS_ADDR_TYPE, &addr);
+	TEST_ASSERT(ret && errno == EINVAL,
+		"ITS region with misaligned address");
+
+	addr = max_phys_size;
+	ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_ITS_ADDR_TYPE, &addr);
+	TEST_ASSERT(ret && errno == E2BIG,
+		"register ITS region with base address beyond IPA range");
+
+	addr = max_phys_size - 0x10000;
+	ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_ITS_ADDR_TYPE, &addr);
+	TEST_ASSERT(ret && errno == E2BIG,
+		"Half of ITS region is beyond IPA range");
+
+	/* This one succeeds setting the ITS base */
+	addr = 0x400000;
+	kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+			    KVM_VGIC_ITS_ADDR_TYPE, &addr);
+
+	addr = 0x300000;
+	ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ADDR,
+				    KVM_VGIC_ITS_ADDR_TYPE, &addr);
+	TEST_ASSERT(ret && errno == EEXIST, "ITS base set again");
+
+	close(its_fd);
+	vm_gic_destroy(&v);
+}
+
+/*
+ * Returns 0 if it's possible to create GIC device of a given type (V2 or V3).
+ */
+int test_kvm_device(uint32_t gic_dev_type)
+{
+	struct kvm_vcpu *vcpus[NR_VCPUS];
+	struct vm_gic v;
+	uint32_t other;
+	int ret;
+
+	v.vm = vm_create_with_vcpus(NR_VCPUS, guest_code, vcpus);
+
+	/* try to create a non existing KVM device */
+	ret = __kvm_test_create_device(v.vm, 0);
+	TEST_ASSERT(ret && errno == ENODEV, "unsupported device");
+
+	/* trial mode */
+	ret = __kvm_test_create_device(v.vm, gic_dev_type);
+	if (ret)
+		return ret;
+	v.gic_fd = kvm_create_device(v.vm, gic_dev_type);
+
+	ret = __kvm_create_device(v.vm, gic_dev_type);
+	TEST_ASSERT(ret < 0 && errno == EEXIST, "create GIC device twice");
+
+	/* try to create the other gic_dev_type */
+	other = VGIC_DEV_IS_V2(gic_dev_type) ? KVM_DEV_TYPE_ARM_VGIC_V3
+					     : KVM_DEV_TYPE_ARM_VGIC_V2;
+
+	if (!__kvm_test_create_device(v.vm, other)) {
+		ret = __kvm_create_device(v.vm, other);
+		TEST_ASSERT(ret < 0 && (errno == EINVAL || errno == EEXIST),
+				"create GIC device while other version exists");
+	}
+
+	vm_gic_destroy(&v);
+
+	return 0;
+}
+
+void run_tests(uint32_t gic_dev_type)
+{
+	test_vcpus_then_vgic(gic_dev_type);
+	test_vgic_then_vcpus(gic_dev_type);
+
+	if (VGIC_DEV_IS_V3(gic_dev_type)) {
+		test_v3_new_redist_regions();
+		test_v3_typer_accesses();
+		test_v3_last_bit_redist_regions();
+		test_v3_last_bit_single_rdist();
+		test_v3_redist_ipa_range_check_at_vcpu_run();
+		test_v3_its_region();
+	}
+}
+
+int main(int ac, char **av)
+{
+	int ret;
+	int pa_bits;
+	int cnt_impl = 0;
+
+	pa_bits = vm_guest_mode_params[VM_MODE_DEFAULT].pa_bits;
+	max_phys_size = 1ULL << pa_bits;
+
+	ret = test_kvm_device(KVM_DEV_TYPE_ARM_VGIC_V3);
+	if (!ret) {
+		pr_info("Running GIC_v3 tests.\n");
+		run_tests(KVM_DEV_TYPE_ARM_VGIC_V3);
+		cnt_impl++;
+	}
+
+	ret = test_kvm_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+	if (!ret) {
+		pr_info("Running GIC_v2 tests.\n");
+		run_tests(KVM_DEV_TYPE_ARM_VGIC_V2);
+		cnt_impl++;
+	}
+
+	if (!cnt_impl) {
+		print_skip("No GICv2 nor GICv3 support");
+		exit(KSFT_SKIP);
+	}
+	return 0;
+}
diff --git a/tools/testing/selftests/kvm/aarch64/vgic_irq.c b/tools/testing/selftests/kvm/aarch64/vgic_irq.c
new file mode 100644
index 0000000000..2e64b4856e
--- /dev/null
+++ b/tools/testing/selftests/kvm/aarch64/vgic_irq.c
@@ -0,0 +1,855 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * vgic_irq.c - Test userspace injection of IRQs
+ *
+ * This test validates the injection of IRQs from userspace using various
+ * methods (e.g., KVM_IRQ_LINE) and modes (e.g., EOI). The guest "asks" the
+ * host to inject a specific intid via a GUEST_SYNC call, and then checks that
+ * it received it.
+ */
+#include <asm/kvm.h>
+#include <asm/kvm_para.h>
+#include <sys/eventfd.h>
+#include <linux/sizes.h>
+
+#include "processor.h"
+#include "test_util.h"
+#include "kvm_util.h"
+#include "gic.h"
+#include "gic_v3.h"
+#include "vgic.h"
+
+#define GICD_BASE_GPA		0x08000000ULL
+#define GICR_BASE_GPA		0x080A0000ULL
+
+/*
+ * Stores the user specified args; it's passed to the guest and to every test
+ * function.
+ */
+struct test_args {
+	uint32_t nr_irqs; /* number of KVM supported IRQs. */
+	bool eoi_split; /* 1 is eoir+dir, 0 is eoir only */
+	bool level_sensitive; /* 1 is level, 0 is edge */
+	int kvm_max_routes; /* output of KVM_CAP_IRQ_ROUTING */
+	bool kvm_supports_irqfd; /* output of KVM_CAP_IRQFD */
+};
+
+/*
+ * KVM implements 32 priority levels:
+ * 0x00 (highest priority) - 0xF8 (lowest priority), in steps of 8
+ *
+ * Note that these macros will still be correct in the case that KVM implements
+ * more priority levels. Also note that 32 is the minimum for GICv3 and GICv2.
+ */
+#define KVM_NUM_PRIOS		32
+#define KVM_PRIO_SHIFT		3 /* steps of 8 = 1 << 3 */
+#define KVM_PRIO_STEPS		(1 << KVM_PRIO_SHIFT) /* 8 */
+#define LOWEST_PRIO		(KVM_NUM_PRIOS - 1)
+#define CPU_PRIO_MASK		(LOWEST_PRIO << KVM_PRIO_SHIFT)	/* 0xf8 */
+#define IRQ_DEFAULT_PRIO	(LOWEST_PRIO - 1)
+#define IRQ_DEFAULT_PRIO_REG	(IRQ_DEFAULT_PRIO << KVM_PRIO_SHIFT) /* 0xf0 */
+
+static void *dist = (void *)GICD_BASE_GPA;
+static void *redist = (void *)GICR_BASE_GPA;
+
+/*
+ * The kvm_inject_* utilities are used by the guest to ask the host to inject
+ * interrupts (e.g., using the KVM_IRQ_LINE ioctl).
+ */
+
+typedef enum {
+	KVM_INJECT_EDGE_IRQ_LINE = 1,
+	KVM_SET_IRQ_LINE,
+	KVM_SET_IRQ_LINE_HIGH,
+	KVM_SET_LEVEL_INFO_HIGH,
+	KVM_INJECT_IRQFD,
+	KVM_WRITE_ISPENDR,
+	KVM_WRITE_ISACTIVER,
+} kvm_inject_cmd;
+
+struct kvm_inject_args {
+	kvm_inject_cmd cmd;
+	uint32_t first_intid;
+	uint32_t num;
+	int level;
+	bool expect_failure;
+};
+
+/* Used on the guest side to perform the hypercall. */
+static void kvm_inject_call(kvm_inject_cmd cmd, uint32_t first_intid,
+		uint32_t num, int level, bool expect_failure);
+
+/* Used on the host side to get the hypercall info. */
+static void kvm_inject_get_call(struct kvm_vm *vm, struct ucall *uc,
+		struct kvm_inject_args *args);
+
+#define _KVM_INJECT_MULTI(cmd, intid, num, expect_failure)			\
+	kvm_inject_call(cmd, intid, num, -1 /* not used */, expect_failure)
+
+#define KVM_INJECT_MULTI(cmd, intid, num)					\
+	_KVM_INJECT_MULTI(cmd, intid, num, false)
+
+#define _KVM_INJECT(cmd, intid, expect_failure)					\
+	_KVM_INJECT_MULTI(cmd, intid, 1, expect_failure)
+
+#define KVM_INJECT(cmd, intid)							\
+	_KVM_INJECT_MULTI(cmd, intid, 1, false)
+
+#define KVM_ACTIVATE(cmd, intid)						\
+	kvm_inject_call(cmd, intid, 1, 1, false);
+
+struct kvm_inject_desc {
+	kvm_inject_cmd cmd;
+	/* can inject PPIs, PPIs, and/or SPIs. */
+	bool sgi, ppi, spi;
+};
+
+static struct kvm_inject_desc inject_edge_fns[] = {
+	/*                                      sgi    ppi    spi */
+	{ KVM_INJECT_EDGE_IRQ_LINE,		false, false, true },
+	{ KVM_INJECT_IRQFD,			false, false, true },
+	{ KVM_WRITE_ISPENDR,			true,  false, true },
+	{ 0, },
+};
+
+static struct kvm_inject_desc inject_level_fns[] = {
+	/*                                      sgi    ppi    spi */
+	{ KVM_SET_IRQ_LINE_HIGH,		false, true,  true },
+	{ KVM_SET_LEVEL_INFO_HIGH,		false, true,  true },
+	{ KVM_INJECT_IRQFD,			false, false, true },
+	{ KVM_WRITE_ISPENDR,			false, true,  true },
+	{ 0, },
+};
+
+static struct kvm_inject_desc set_active_fns[] = {
+	/*                                      sgi    ppi    spi */
+	{ KVM_WRITE_ISACTIVER,			true,  true,  true },
+	{ 0, },
+};
+
+#define for_each_inject_fn(t, f)						\
+	for ((f) = (t); (f)->cmd; (f)++)
+
+#define for_each_supported_inject_fn(args, t, f)				\
+	for_each_inject_fn(t, f)						\
+		if ((args)->kvm_supports_irqfd || (f)->cmd != KVM_INJECT_IRQFD)
+
+#define for_each_supported_activate_fn(args, t, f)				\
+	for_each_supported_inject_fn((args), (t), (f))
+
+/* Shared between the guest main thread and the IRQ handlers. */
+volatile uint64_t irq_handled;
+volatile uint32_t irqnr_received[MAX_SPI + 1];
+
+static void reset_stats(void)
+{
+	int i;
+
+	irq_handled = 0;
+	for (i = 0; i <= MAX_SPI; i++)
+		irqnr_received[i] = 0;
+}
+
+static uint64_t gic_read_ap1r0(void)
+{
+	uint64_t reg = read_sysreg_s(SYS_ICV_AP1R0_EL1);
+
+	dsb(sy);
+	return reg;
+}
+
+static void gic_write_ap1r0(uint64_t val)
+{
+	write_sysreg_s(val, SYS_ICV_AP1R0_EL1);
+	isb();
+}
+
+static void guest_set_irq_line(uint32_t intid, uint32_t level);
+
+static void guest_irq_generic_handler(bool eoi_split, bool level_sensitive)
+{
+	uint32_t intid = gic_get_and_ack_irq();
+
+	if (intid == IAR_SPURIOUS)
+		return;
+
+	GUEST_ASSERT(gic_irq_get_active(intid));
+
+	if (!level_sensitive)
+		GUEST_ASSERT(!gic_irq_get_pending(intid));
+
+	if (level_sensitive)
+		guest_set_irq_line(intid, 0);
+
+	GUEST_ASSERT(intid < MAX_SPI);
+	irqnr_received[intid] += 1;
+	irq_handled += 1;
+
+	gic_set_eoi(intid);
+	GUEST_ASSERT_EQ(gic_read_ap1r0(), 0);
+	if (eoi_split)
+		gic_set_dir(intid);
+
+	GUEST_ASSERT(!gic_irq_get_active(intid));
+	GUEST_ASSERT(!gic_irq_get_pending(intid));
+}
+
+static void kvm_inject_call(kvm_inject_cmd cmd, uint32_t first_intid,
+		uint32_t num, int level, bool expect_failure)
+{
+	struct kvm_inject_args args = {
+		.cmd = cmd,
+		.first_intid = first_intid,
+		.num = num,
+		.level = level,
+		.expect_failure = expect_failure,
+	};
+	GUEST_SYNC(&args);
+}
+
+#define GUEST_ASSERT_IAR_EMPTY()						\
+do { 										\
+	uint32_t _intid;							\
+	_intid = gic_get_and_ack_irq();						\
+	GUEST_ASSERT(_intid == 0 || _intid == IAR_SPURIOUS);			\
+} while (0)
+
+#define CAT_HELPER(a, b) a ## b
+#define CAT(a, b) CAT_HELPER(a, b)
+#define PREFIX guest_irq_handler_
+#define GUEST_IRQ_HANDLER_NAME(split, lev) CAT(PREFIX, CAT(split, lev))
+#define GENERATE_GUEST_IRQ_HANDLER(split, lev)					\
+static void CAT(PREFIX, CAT(split, lev))(struct ex_regs *regs)			\
+{										\
+	guest_irq_generic_handler(split, lev);					\
+}
+
+GENERATE_GUEST_IRQ_HANDLER(0, 0);
+GENERATE_GUEST_IRQ_HANDLER(0, 1);
+GENERATE_GUEST_IRQ_HANDLER(1, 0);
+GENERATE_GUEST_IRQ_HANDLER(1, 1);
+
+static void (*guest_irq_handlers[2][2])(struct ex_regs *) = {
+	{GUEST_IRQ_HANDLER_NAME(0, 0), GUEST_IRQ_HANDLER_NAME(0, 1),},
+	{GUEST_IRQ_HANDLER_NAME(1, 0), GUEST_IRQ_HANDLER_NAME(1, 1),},
+};
+
+static void reset_priorities(struct test_args *args)
+{
+	int i;
+
+	for (i = 0; i < args->nr_irqs; i++)
+		gic_set_priority(i, IRQ_DEFAULT_PRIO_REG);
+}
+
+static void guest_set_irq_line(uint32_t intid, uint32_t level)
+{
+	kvm_inject_call(KVM_SET_IRQ_LINE, intid, 1, level, false);
+}
+
+static void test_inject_fail(struct test_args *args,
+		uint32_t intid, kvm_inject_cmd cmd)
+{
+	reset_stats();
+
+	_KVM_INJECT(cmd, intid, true);
+	/* no IRQ to handle on entry */
+
+	GUEST_ASSERT_EQ(irq_handled, 0);
+	GUEST_ASSERT_IAR_EMPTY();
+}
+
+static void guest_inject(struct test_args *args,
+		uint32_t first_intid, uint32_t num,
+		kvm_inject_cmd cmd)
+{
+	uint32_t i;
+
+	reset_stats();
+
+	/* Cycle over all priorities to make things more interesting. */
+	for (i = first_intid; i < num + first_intid; i++)
+		gic_set_priority(i, (i % (KVM_NUM_PRIOS - 1)) << 3);
+
+	asm volatile("msr daifset, #2" : : : "memory");
+	KVM_INJECT_MULTI(cmd, first_intid, num);
+
+	while (irq_handled < num) {
+		asm volatile("wfi\n"
+			     "msr daifclr, #2\n"
+			     /* handle IRQ */
+			     "msr daifset, #2\n"
+			     : : : "memory");
+	}
+	asm volatile("msr daifclr, #2" : : : "memory");
+
+	GUEST_ASSERT_EQ(irq_handled, num);
+	for (i = first_intid; i < num + first_intid; i++)
+		GUEST_ASSERT_EQ(irqnr_received[i], 1);
+	GUEST_ASSERT_IAR_EMPTY();
+
+	reset_priorities(args);
+}
+
+/*
+ * Restore the active state of multiple concurrent IRQs (given by
+ * concurrent_irqs).  This does what a live-migration would do on the
+ * destination side assuming there are some active IRQs that were not
+ * deactivated yet.
+ */
+static void guest_restore_active(struct test_args *args,
+		uint32_t first_intid, uint32_t num,
+		kvm_inject_cmd cmd)
+{
+	uint32_t prio, intid, ap1r;
+	int i;
+
+	/*
+	 * Set the priorities of the first (KVM_NUM_PRIOS - 1) IRQs
+	 * in descending order, so intid+1 can preempt intid.
+	 */
+	for (i = 0, prio = (num - 1) * 8; i < num; i++, prio -= 8) {
+		GUEST_ASSERT(prio >= 0);
+		intid = i + first_intid;
+		gic_set_priority(intid, prio);
+	}
+
+	/*
+	 * In a real migration, KVM would restore all GIC state before running
+	 * guest code.
+	 */
+	for (i = 0; i < num; i++) {
+		intid = i + first_intid;
+		KVM_ACTIVATE(cmd, intid);
+		ap1r = gic_read_ap1r0();
+		ap1r |= 1U << i;
+		gic_write_ap1r0(ap1r);
+	}
+
+	/* This is where the "migration" would occur. */
+
+	/* finish handling the IRQs starting with the highest priority one. */
+	for (i = 0; i < num; i++) {
+		intid = num - i - 1 + first_intid;
+		gic_set_eoi(intid);
+		if (args->eoi_split)
+			gic_set_dir(intid);
+	}
+
+	for (i = 0; i < num; i++)
+		GUEST_ASSERT(!gic_irq_get_active(i + first_intid));
+	GUEST_ASSERT_EQ(gic_read_ap1r0(), 0);
+	GUEST_ASSERT_IAR_EMPTY();
+}
+
+/*
+ * Polls the IAR until it's not a spurious interrupt.
+ *
+ * This function should only be used in test_inject_preemption (with IRQs
+ * masked).
+ */
+static uint32_t wait_for_and_activate_irq(void)
+{
+	uint32_t intid;
+
+	do {
+		asm volatile("wfi" : : : "memory");
+		intid = gic_get_and_ack_irq();
+	} while (intid == IAR_SPURIOUS);
+
+	return intid;
+}
+
+/*
+ * Inject multiple concurrent IRQs (num IRQs starting at first_intid) and
+ * handle them without handling the actual exceptions.  This is done by masking
+ * interrupts for the whole test.
+ */
+static void test_inject_preemption(struct test_args *args,
+		uint32_t first_intid, int num,
+		kvm_inject_cmd cmd)
+{
+	uint32_t intid, prio, step = KVM_PRIO_STEPS;
+	int i;
+
+	/* Set the priorities of the first (KVM_NUM_PRIOS - 1) IRQs
+	 * in descending order, so intid+1 can preempt intid.
+	 */
+	for (i = 0, prio = (num - 1) * step; i < num; i++, prio -= step) {
+		GUEST_ASSERT(prio >= 0);
+		intid = i + first_intid;
+		gic_set_priority(intid, prio);
+	}
+
+	local_irq_disable();
+
+	for (i = 0; i < num; i++) {
+		uint32_t tmp;
+		intid = i + first_intid;
+		KVM_INJECT(cmd, intid);
+		/* Each successive IRQ will preempt the previous one. */
+		tmp = wait_for_and_activate_irq();
+		GUEST_ASSERT_EQ(tmp, intid);
+		if (args->level_sensitive)
+			guest_set_irq_line(intid, 0);
+	}
+
+	/* finish handling the IRQs starting with the highest priority one. */
+	for (i = 0; i < num; i++) {
+		intid = num - i - 1 + first_intid;
+		gic_set_eoi(intid);
+		if (args->eoi_split)
+			gic_set_dir(intid);
+	}
+
+	local_irq_enable();
+
+	for (i = 0; i < num; i++)
+		GUEST_ASSERT(!gic_irq_get_active(i + first_intid));
+	GUEST_ASSERT_EQ(gic_read_ap1r0(), 0);
+	GUEST_ASSERT_IAR_EMPTY();
+
+	reset_priorities(args);
+}
+
+static void test_injection(struct test_args *args, struct kvm_inject_desc *f)
+{
+	uint32_t nr_irqs = args->nr_irqs;
+
+	if (f->sgi) {
+		guest_inject(args, MIN_SGI, 1, f->cmd);
+		guest_inject(args, 0, 16, f->cmd);
+	}
+
+	if (f->ppi)
+		guest_inject(args, MIN_PPI, 1, f->cmd);
+
+	if (f->spi) {
+		guest_inject(args, MIN_SPI, 1, f->cmd);
+		guest_inject(args, nr_irqs - 1, 1, f->cmd);
+		guest_inject(args, MIN_SPI, nr_irqs - MIN_SPI, f->cmd);
+	}
+}
+
+static void test_injection_failure(struct test_args *args,
+		struct kvm_inject_desc *f)
+{
+	uint32_t bad_intid[] = { args->nr_irqs, 1020, 1024, 1120, 5120, ~0U, };
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(bad_intid); i++)
+		test_inject_fail(args, bad_intid[i], f->cmd);
+}
+
+static void test_preemption(struct test_args *args, struct kvm_inject_desc *f)
+{
+	/*
+	 * Test up to 4 levels of preemption. The reason is that KVM doesn't
+	 * currently implement the ability to have more than the number-of-LRs
+	 * number of concurrently active IRQs. The number of LRs implemented is
+	 * IMPLEMENTATION DEFINED, however, it seems that most implement 4.
+	 */
+	if (f->sgi)
+		test_inject_preemption(args, MIN_SGI, 4, f->cmd);
+
+	if (f->ppi)
+		test_inject_preemption(args, MIN_PPI, 4, f->cmd);
+
+	if (f->spi)
+		test_inject_preemption(args, MIN_SPI, 4, f->cmd);
+}
+
+static void test_restore_active(struct test_args *args, struct kvm_inject_desc *f)
+{
+	/* Test up to 4 active IRQs. Same reason as in test_preemption. */
+	if (f->sgi)
+		guest_restore_active(args, MIN_SGI, 4, f->cmd);
+
+	if (f->ppi)
+		guest_restore_active(args, MIN_PPI, 4, f->cmd);
+
+	if (f->spi)
+		guest_restore_active(args, MIN_SPI, 4, f->cmd);
+}
+
+static void guest_code(struct test_args *args)
+{
+	uint32_t i, nr_irqs = args->nr_irqs;
+	bool level_sensitive = args->level_sensitive;
+	struct kvm_inject_desc *f, *inject_fns;
+
+	gic_init(GIC_V3, 1, dist, redist);
+
+	for (i = 0; i < nr_irqs; i++)
+		gic_irq_enable(i);
+
+	for (i = MIN_SPI; i < nr_irqs; i++)
+		gic_irq_set_config(i, !level_sensitive);
+
+	gic_set_eoi_split(args->eoi_split);
+
+	reset_priorities(args);
+	gic_set_priority_mask(CPU_PRIO_MASK);
+
+	inject_fns  = level_sensitive ? inject_level_fns
+				      : inject_edge_fns;
+
+	local_irq_enable();
+
+	/* Start the tests. */
+	for_each_supported_inject_fn(args, inject_fns, f) {
+		test_injection(args, f);
+		test_preemption(args, f);
+		test_injection_failure(args, f);
+	}
+
+	/*
+	 * Restore the active state of IRQs. This would happen when live
+	 * migrating IRQs in the middle of being handled.
+	 */
+	for_each_supported_activate_fn(args, set_active_fns, f)
+		test_restore_active(args, f);
+
+	GUEST_DONE();
+}
+
+static void kvm_irq_line_check(struct kvm_vm *vm, uint32_t intid, int level,
+			struct test_args *test_args, bool expect_failure)
+{
+	int ret;
+
+	if (!expect_failure) {
+		kvm_arm_irq_line(vm, intid, level);
+	} else {
+		/* The interface doesn't allow larger intid's. */
+		if (intid > KVM_ARM_IRQ_NUM_MASK)
+			return;
+
+		ret = _kvm_arm_irq_line(vm, intid, level);
+		TEST_ASSERT(ret != 0 && errno == EINVAL,
+				"Bad intid %i did not cause KVM_IRQ_LINE "
+				"error: rc: %i errno: %i", intid, ret, errno);
+	}
+}
+
+void kvm_irq_set_level_info_check(int gic_fd, uint32_t intid, int level,
+			bool expect_failure)
+{
+	if (!expect_failure) {
+		kvm_irq_set_level_info(gic_fd, intid, level);
+	} else {
+		int ret = _kvm_irq_set_level_info(gic_fd, intid, level);
+		/*
+		 * The kernel silently fails for invalid SPIs and SGIs (which
+		 * are not level-sensitive). It only checks for intid to not
+		 * spill over 1U << 10 (the max reserved SPI). Also, callers
+		 * are supposed to mask the intid with 0x3ff (1023).
+		 */
+		if (intid > VGIC_MAX_RESERVED)
+			TEST_ASSERT(ret != 0 && errno == EINVAL,
+				"Bad intid %i did not cause VGIC_GRP_LEVEL_INFO "
+				"error: rc: %i errno: %i", intid, ret, errno);
+		else
+			TEST_ASSERT(!ret, "KVM_DEV_ARM_VGIC_GRP_LEVEL_INFO "
+				"for intid %i failed, rc: %i errno: %i",
+				intid, ret, errno);
+	}
+}
+
+static void kvm_set_gsi_routing_irqchip_check(struct kvm_vm *vm,
+		uint32_t intid, uint32_t num, uint32_t kvm_max_routes,
+		bool expect_failure)
+{
+	struct kvm_irq_routing *routing;
+	int ret;
+	uint64_t i;
+
+	assert(num <= kvm_max_routes && kvm_max_routes <= KVM_MAX_IRQ_ROUTES);
+
+	routing = kvm_gsi_routing_create();
+	for (i = intid; i < (uint64_t)intid + num; i++)
+		kvm_gsi_routing_irqchip_add(routing, i - MIN_SPI, i - MIN_SPI);
+
+	if (!expect_failure) {
+		kvm_gsi_routing_write(vm, routing);
+	} else {
+		ret = _kvm_gsi_routing_write(vm, routing);
+		/* The kernel only checks e->irqchip.pin >= KVM_IRQCHIP_NUM_PINS */
+		if (((uint64_t)intid + num - 1 - MIN_SPI) >= KVM_IRQCHIP_NUM_PINS)
+			TEST_ASSERT(ret != 0 && errno == EINVAL,
+				"Bad intid %u did not cause KVM_SET_GSI_ROUTING "
+				"error: rc: %i errno: %i", intid, ret, errno);
+		else
+			TEST_ASSERT(ret == 0, "KVM_SET_GSI_ROUTING "
+				"for intid %i failed, rc: %i errno: %i",
+				intid, ret, errno);
+	}
+}
+
+static void kvm_irq_write_ispendr_check(int gic_fd, uint32_t intid,
+					struct kvm_vcpu *vcpu,
+					bool expect_failure)
+{
+	/*
+	 * Ignore this when expecting failure as invalid intids will lead to
+	 * either trying to inject SGIs when we configured the test to be
+	 * level_sensitive (or the reverse), or inject large intids which
+	 * will lead to writing above the ISPENDR register space (and we
+	 * don't want to do that either).
+	 */
+	if (!expect_failure)
+		kvm_irq_write_ispendr(gic_fd, intid, vcpu);
+}
+
+static void kvm_routing_and_irqfd_check(struct kvm_vm *vm,
+		uint32_t intid, uint32_t num, uint32_t kvm_max_routes,
+		bool expect_failure)
+{
+	int fd[MAX_SPI];
+	uint64_t val;
+	int ret, f;
+	uint64_t i;
+
+	/*
+	 * There is no way to try injecting an SGI or PPI as the interface
+	 * starts counting from the first SPI (above the private ones), so just
+	 * exit.
+	 */
+	if (INTID_IS_SGI(intid) || INTID_IS_PPI(intid))
+		return;
+
+	kvm_set_gsi_routing_irqchip_check(vm, intid, num,
+			kvm_max_routes, expect_failure);
+
+	/*
+	 * If expect_failure, then just to inject anyway. These
+	 * will silently fail. And in any case, the guest will check
+	 * that no actual interrupt was injected for those cases.
+	 */
+
+	for (f = 0, i = intid; i < (uint64_t)intid + num; i++, f++) {
+		fd[f] = eventfd(0, 0);
+		TEST_ASSERT(fd[f] != -1, __KVM_SYSCALL_ERROR("eventfd()", fd[f]));
+	}
+
+	for (f = 0, i = intid; i < (uint64_t)intid + num; i++, f++) {
+		struct kvm_irqfd irqfd = {
+			.fd  = fd[f],
+			.gsi = i - MIN_SPI,
+		};
+		assert(i <= (uint64_t)UINT_MAX);
+		vm_ioctl(vm, KVM_IRQFD, &irqfd);
+	}
+
+	for (f = 0, i = intid; i < (uint64_t)intid + num; i++, f++) {
+		val = 1;
+		ret = write(fd[f], &val, sizeof(uint64_t));
+		TEST_ASSERT(ret == sizeof(uint64_t),
+			    __KVM_SYSCALL_ERROR("write()", ret));
+	}
+
+	for (f = 0, i = intid; i < (uint64_t)intid + num; i++, f++)
+		close(fd[f]);
+}
+
+/* handles the valid case: intid=0xffffffff num=1 */
+#define for_each_intid(first, num, tmp, i)					\
+	for ((tmp) = (i) = (first);						\
+		(tmp) < (uint64_t)(first) + (uint64_t)(num);			\
+		(tmp)++, (i)++)
+
+static void run_guest_cmd(struct kvm_vcpu *vcpu, int gic_fd,
+			  struct kvm_inject_args *inject_args,
+			  struct test_args *test_args)
+{
+	kvm_inject_cmd cmd = inject_args->cmd;
+	uint32_t intid = inject_args->first_intid;
+	uint32_t num = inject_args->num;
+	int level = inject_args->level;
+	bool expect_failure = inject_args->expect_failure;
+	struct kvm_vm *vm = vcpu->vm;
+	uint64_t tmp;
+	uint32_t i;
+
+	/* handles the valid case: intid=0xffffffff num=1 */
+	assert(intid < UINT_MAX - num || num == 1);
+
+	switch (cmd) {
+	case KVM_INJECT_EDGE_IRQ_LINE:
+		for_each_intid(intid, num, tmp, i)
+			kvm_irq_line_check(vm, i, 1, test_args,
+					expect_failure);
+		for_each_intid(intid, num, tmp, i)
+			kvm_irq_line_check(vm, i, 0, test_args,
+					expect_failure);
+		break;
+	case KVM_SET_IRQ_LINE:
+		for_each_intid(intid, num, tmp, i)
+			kvm_irq_line_check(vm, i, level, test_args,
+					expect_failure);
+		break;
+	case KVM_SET_IRQ_LINE_HIGH:
+		for_each_intid(intid, num, tmp, i)
+			kvm_irq_line_check(vm, i, 1, test_args,
+					expect_failure);
+		break;
+	case KVM_SET_LEVEL_INFO_HIGH:
+		for_each_intid(intid, num, tmp, i)
+			kvm_irq_set_level_info_check(gic_fd, i, 1,
+					expect_failure);
+		break;
+	case KVM_INJECT_IRQFD:
+		kvm_routing_and_irqfd_check(vm, intid, num,
+					test_args->kvm_max_routes,
+					expect_failure);
+		break;
+	case KVM_WRITE_ISPENDR:
+		for (i = intid; i < intid + num; i++)
+			kvm_irq_write_ispendr_check(gic_fd, i, vcpu,
+						    expect_failure);
+		break;
+	case KVM_WRITE_ISACTIVER:
+		for (i = intid; i < intid + num; i++)
+			kvm_irq_write_isactiver(gic_fd, i, vcpu);
+		break;
+	default:
+		break;
+	}
+}
+
+static void kvm_inject_get_call(struct kvm_vm *vm, struct ucall *uc,
+		struct kvm_inject_args *args)
+{
+	struct kvm_inject_args *kvm_args_hva;
+	vm_vaddr_t kvm_args_gva;
+
+	kvm_args_gva = uc->args[1];
+	kvm_args_hva = (struct kvm_inject_args *)addr_gva2hva(vm, kvm_args_gva);
+	memcpy(args, kvm_args_hva, sizeof(struct kvm_inject_args));
+}
+
+static void print_args(struct test_args *args)
+{
+	printf("nr-irqs=%d level-sensitive=%d eoi-split=%d\n",
+			args->nr_irqs, args->level_sensitive,
+			args->eoi_split);
+}
+
+static void test_vgic(uint32_t nr_irqs, bool level_sensitive, bool eoi_split)
+{
+	struct ucall uc;
+	int gic_fd;
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	struct kvm_inject_args inject_args;
+	vm_vaddr_t args_gva;
+
+	struct test_args args = {
+		.nr_irqs = nr_irqs,
+		.level_sensitive = level_sensitive,
+		.eoi_split = eoi_split,
+		.kvm_max_routes = kvm_check_cap(KVM_CAP_IRQ_ROUTING),
+		.kvm_supports_irqfd = kvm_check_cap(KVM_CAP_IRQFD),
+	};
+
+	print_args(&args);
+
+	vm = vm_create_with_one_vcpu(&vcpu, guest_code);
+
+	vm_init_descriptor_tables(vm);
+	vcpu_init_descriptor_tables(vcpu);
+
+	/* Setup the guest args page (so it gets the args). */
+	args_gva = vm_vaddr_alloc_page(vm);
+	memcpy(addr_gva2hva(vm, args_gva), &args, sizeof(args));
+	vcpu_args_set(vcpu, 1, args_gva);
+
+	gic_fd = vgic_v3_setup(vm, 1, nr_irqs,
+			GICD_BASE_GPA, GICR_BASE_GPA);
+	__TEST_REQUIRE(gic_fd >= 0, "Failed to create vgic-v3, skipping");
+
+	vm_install_exception_handler(vm, VECTOR_IRQ_CURRENT,
+		guest_irq_handlers[args.eoi_split][args.level_sensitive]);
+
+	while (1) {
+		vcpu_run(vcpu);
+
+		switch (get_ucall(vcpu, &uc)) {
+		case UCALL_SYNC:
+			kvm_inject_get_call(vm, &uc, &inject_args);
+			run_guest_cmd(vcpu, gic_fd, &inject_args, &args);
+			break;
+		case UCALL_ABORT:
+			REPORT_GUEST_ASSERT(uc);
+			break;
+		case UCALL_DONE:
+			goto done;
+		default:
+			TEST_FAIL("Unknown ucall %lu", uc.cmd);
+		}
+	}
+
+done:
+	close(gic_fd);
+	kvm_vm_free(vm);
+}
+
+static void help(const char *name)
+{
+	printf(
+	"\n"
+	"usage: %s [-n num_irqs] [-e eoi_split] [-l level_sensitive]\n", name);
+	printf(" -n: specify number of IRQs to setup the vgic with. "
+		"It has to be a multiple of 32 and between 64 and 1024.\n");
+	printf(" -e: if 1 then EOI is split into a write to DIR on top "
+		"of writing EOI.\n");
+	printf(" -l: specify whether the IRQs are level-sensitive (1) or not (0).");
+	puts("");
+	exit(1);
+}
+
+int main(int argc, char **argv)
+{
+	uint32_t nr_irqs = 64;
+	bool default_args = true;
+	bool level_sensitive = false;
+	int opt;
+	bool eoi_split = false;
+
+	while ((opt = getopt(argc, argv, "hn:e:l:")) != -1) {
+		switch (opt) {
+		case 'n':
+			nr_irqs = atoi_non_negative("Number of IRQs", optarg);
+			if (nr_irqs > 1024 || nr_irqs % 32)
+				help(argv[0]);
+			break;
+		case 'e':
+			eoi_split = (bool)atoi_paranoid(optarg);
+			default_args = false;
+			break;
+		case 'l':
+			level_sensitive = (bool)atoi_paranoid(optarg);
+			default_args = false;
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			break;
+		}
+	}
+
+	/*
+	 * If the user just specified nr_irqs and/or gic_version, then run all
+	 * combinations.
+	 */
+	if (default_args) {
+		test_vgic(nr_irqs, false /* level */, false /* eoi_split */);
+		test_vgic(nr_irqs, false /* level */, true /* eoi_split */);
+		test_vgic(nr_irqs, true /* level */, false /* eoi_split */);
+		test_vgic(nr_irqs, true /* level */, true /* eoi_split */);
+	} else {
+		test_vgic(nr_irqs, level_sensitive, eoi_split);
+	}
+
+	return 0;
+}