KVM: selftests: Stop conflating vCPU index and ID in perf tests

Track vCPUs by their 'struct kvm_vcpu' object, and stop assuming that a
vCPU's ID is the same as its index when referencing a vCPU's metadata.

Signed-off-by: Sean Christopherson <[e-mail]>
Signed-off-by: Paolo Bonzini <[e-mail]>

tools/testing/selftests/kvm/access_tracking_perf_test.c

···
 	uint64_t vcpu_memory_bytes;
 
 	/* The number of vCPUs to create in the VM. */
-	int vcpus;
+	int nr_vcpus;
 };
 
 static uint64_t pread_uint64(int fd, const char *filename, uint64_t index)
···
 		    "Set page_idle bits for PFN 0x%" PRIx64, pfn);
 }
 
-static void mark_vcpu_memory_idle(struct kvm_vm *vm, int vcpu_id)
+static void mark_vcpu_memory_idle(struct kvm_vm *vm,
+				  struct perf_test_vcpu_args *vcpu_args)
 {
-	uint64_t base_gva = perf_test_args.vcpu_args[vcpu_id].gva;
-	uint64_t pages = perf_test_args.vcpu_args[vcpu_id].pages;
+	int vcpu_idx = vcpu_args->vcpu_idx;
+	uint64_t base_gva = vcpu_args->gva;
+	uint64_t pages = vcpu_args->pages;
 	uint64_t page;
 	uint64_t still_idle = 0;
 	uint64_t no_pfn = 0;
···
 	int pagemap_fd;
 
 	/* If vCPUs are using an overlapping region, let vCPU 0 mark it idle. */
-	if (overlap_memory_access && vcpu_id)
+	if (overlap_memory_access && vcpu_idx)
 		return;
 
 	page_idle_fd = open("/sys/kernel/mm/page_idle/bitmap", O_RDWR);
···
 	 */
 	TEST_ASSERT(no_pfn < pages / 100,
 		    "vCPU %d: No PFN for %" PRIu64 " out of %" PRIu64 " pages.",
-		    vcpu_id, no_pfn, pages);
+		    vcpu_idx, no_pfn, pages);
 
 	/*
 	 * Test that at least 90% of memory has been marked idle (the rest might
···
 	TEST_ASSERT(still_idle < pages / 10,
 		    "vCPU%d: Too many pages still idle (%"PRIu64 " out of %"
 		    PRIu64 ").\n",
-		    vcpu_id, still_idle, pages);
+		    vcpu_idx, still_idle, pages);
 
 	close(page_idle_fd);
 	close(pagemap_fd);
 }
 
-static void assert_ucall(struct kvm_vm *vm, uint32_t vcpu_id,
-			 uint64_t expected_ucall)
+static void assert_ucall(struct kvm_vcpu *vcpu, uint64_t expected_ucall)
 {
 	struct ucall uc;
-	uint64_t actual_ucall = get_ucall(vm, vcpu_id, &uc);
+	uint64_t actual_ucall = get_ucall(vcpu->vm, vcpu->id, &uc);
 
 	TEST_ASSERT(expected_ucall == actual_ucall,
 		    "Guest exited unexpectedly (expected ucall %" PRIu64
···
 
 static void vcpu_thread_main(struct perf_test_vcpu_args *vcpu_args)
 {
+	struct kvm_vcpu *vcpu = vcpu_args->vcpu;
 	struct kvm_vm *vm = perf_test_args.vm;
-	int vcpu_id = vcpu_args->vcpu_id;
+	int vcpu_idx = vcpu_args->vcpu_idx;
 	int current_iteration = 0;
 
 	while (spin_wait_for_next_iteration(&current_iteration)) {
 		switch (READ_ONCE(iteration_work)) {
 		case ITERATION_ACCESS_MEMORY:
-			vcpu_run(vm, vcpu_id);
-			assert_ucall(vm, vcpu_id, UCALL_SYNC);
+			vcpu_run(vm, vcpu->id);
+			assert_ucall(vcpu, UCALL_SYNC);
 			break;
 		case ITERATION_MARK_IDLE:
-			mark_vcpu_memory_idle(vm, vcpu_id);
+			mark_vcpu_memory_idle(vm, vcpu_args);
 			break;
 		};
 
-		vcpu_last_completed_iteration[vcpu_id] = current_iteration;
+		vcpu_last_completed_iteration[vcpu_idx] = current_iteration;
 	}
 }
 
-static void spin_wait_for_vcpu(int vcpu_id, int target_iteration)
+static void spin_wait_for_vcpu(int vcpu_idx, int target_iteration)
 {
-	while (READ_ONCE(vcpu_last_completed_iteration[vcpu_id]) !=
+	while (READ_ONCE(vcpu_last_completed_iteration[vcpu_idx]) !=
 	       target_iteration) {
 		continue;
 	}
···
 	ACCESS_WRITE,
 };
 
-static void run_iteration(struct kvm_vm *vm, int vcpus, const char *description)
+static void run_iteration(struct kvm_vm *vm, int nr_vcpus, const char *description)
 {
 	struct timespec ts_start;
 	struct timespec ts_elapsed;
-	int next_iteration;
-	int vcpu_id;
+	int next_iteration, i;
 
 	/* Kick off the vCPUs by incrementing iteration. */
 	next_iteration = ++iteration;
···
 	clock_gettime(CLOCK_MONOTONIC, &ts_start);
 
 	/* Wait for all vCPUs to finish the iteration. */
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++)
-		spin_wait_for_vcpu(vcpu_id, next_iteration);
+	for (i = 0; i < nr_vcpus; i++)
+		spin_wait_for_vcpu(i, next_iteration);
 
 	ts_elapsed = timespec_elapsed(ts_start);
 	pr_info("%-30s: %ld.%09lds\n",
 		description, ts_elapsed.tv_sec, ts_elapsed.tv_nsec);
 }
 
-static void access_memory(struct kvm_vm *vm, int vcpus, enum access_type access,
-			  const char *description)
+static void access_memory(struct kvm_vm *vm, int nr_vcpus,
+			  enum access_type access, const char *description)
 {
 	perf_test_set_wr_fract(vm, (access == ACCESS_READ) ? INT_MAX : 1);
 	iteration_work = ITERATION_ACCESS_MEMORY;
-	run_iteration(vm, vcpus, description);
+	run_iteration(vm, nr_vcpus, description);
 }
 
-static void mark_memory_idle(struct kvm_vm *vm, int vcpus)
+static void mark_memory_idle(struct kvm_vm *vm, int nr_vcpus)
 {
 	/*
 	 * Even though this parallelizes the work across vCPUs, this is still a
···
 	 */
 	pr_debug("Marking VM memory idle (slow)...\n");
 	iteration_work = ITERATION_MARK_IDLE;
-	run_iteration(vm, vcpus, "Mark memory idle");
+	run_iteration(vm, nr_vcpus, "Mark memory idle");
 }
 
 static void run_test(enum vm_guest_mode mode, void *arg)
 {
 	struct test_params *params = arg;
 	struct kvm_vm *vm;
-	int vcpus = params->vcpus;
+	int nr_vcpus = params->nr_vcpus;
 
-	vm = perf_test_create_vm(mode, vcpus, params->vcpu_memory_bytes, 1,
+	vm = perf_test_create_vm(mode, nr_vcpus, params->vcpu_memory_bytes, 1,
 				 params->backing_src, !overlap_memory_access);
 
-	perf_test_start_vcpu_threads(vcpus, vcpu_thread_main);
+	perf_test_start_vcpu_threads(nr_vcpus, vcpu_thread_main);
 
 	pr_info("\n");
-	access_memory(vm, vcpus, ACCESS_WRITE, "Populating memory");
+	access_memory(vm, nr_vcpus, ACCESS_WRITE, "Populating memory");
 
 	/* As a control, read and write to the populated memory first. */
-	access_memory(vm, vcpus, ACCESS_WRITE, "Writing to populated memory");
-	access_memory(vm, vcpus, ACCESS_READ, "Reading from populated memory");
+	access_memory(vm, nr_vcpus, ACCESS_WRITE, "Writing to populated memory");
+	access_memory(vm, nr_vcpus, ACCESS_READ, "Reading from populated memory");
 
 	/* Repeat on memory that has been marked as idle. */
-	mark_memory_idle(vm, vcpus);
-	access_memory(vm, vcpus, ACCESS_WRITE, "Writing to idle memory");
-	mark_memory_idle(vm, vcpus);
-	access_memory(vm, vcpus, ACCESS_READ, "Reading from idle memory");
+	mark_memory_idle(vm, nr_vcpus);
+	access_memory(vm, nr_vcpus, ACCESS_WRITE, "Writing to idle memory");
+	mark_memory_idle(vm, nr_vcpus);
+	access_memory(vm, nr_vcpus, ACCESS_READ, "Reading from idle memory");
 
 	/* Set done to signal the vCPU threads to exit */
 	done = true;
 
-	perf_test_join_vcpu_threads(vcpus);
+	perf_test_join_vcpu_threads(nr_vcpus);
 	perf_test_destroy_vm(vm);
 }
 
···
 	struct test_params params = {
 		.backing_src = DEFAULT_VM_MEM_SRC,
 		.vcpu_memory_bytes = DEFAULT_PER_VCPU_MEM_SIZE,
-		.vcpus = 1,
+		.nr_vcpus = 1,
 	};
 	int page_idle_fd;
 	int opt;
···
 			params.vcpu_memory_bytes = parse_size(optarg);
 			break;
 		case 'v':
-			params.vcpus = atoi(optarg);
+			params.nr_vcpus = atoi(optarg);
 			break;
 		case 'o':
 			overlap_memory_access = true;

tools/testing/selftests/kvm/demand_paging_test.c

···
 
 static void vcpu_worker(struct perf_test_vcpu_args *vcpu_args)
 {
-	int ret;
-	int vcpu_id = vcpu_args->vcpu_id;
+	struct kvm_vcpu *vcpu = vcpu_args->vcpu;
 	struct kvm_vm *vm = perf_test_args.vm;
-	struct kvm_run *run;
+	int vcpu_idx = vcpu_args->vcpu_idx;
+	struct kvm_run *run = vcpu->run;
 	struct timespec start;
 	struct timespec ts_diff;
-
-	run = vcpu_state(vm, vcpu_id);
+	int ret;
 
 	clock_gettime(CLOCK_MONOTONIC, &start);
 
 	/* Let the guest access its memory */
-	ret = _vcpu_run(vm, vcpu_id);
+	ret = _vcpu_run(vm, vcpu->id);
 	TEST_ASSERT(ret == 0, "vcpu_run failed: %d\n", ret);
-	if (get_ucall(vm, vcpu_id, NULL) != UCALL_SYNC) {
+	if (get_ucall(vm, vcpu->id, NULL) != UCALL_SYNC) {
 		TEST_ASSERT(false,
 			    "Invalid guest sync status: exit_reason=%s\n",
 			    exit_reason_str(run->exit_reason));
 	}
 
 	ts_diff = timespec_elapsed(start);
-	PER_VCPU_DEBUG("vCPU %d execution time: %ld.%.9lds\n", vcpu_id,
+	PER_VCPU_DEBUG("vCPU %d execution time: %ld.%.9lds\n", vcpu_idx,
 		       ts_diff.tv_sec, ts_diff.tv_nsec);
 }
 
···
 	struct timespec ts_diff;
 	int *pipefds = NULL;
 	struct kvm_vm *vm;
-	int vcpu_id;
-	int r;
+	int r, i;
 
 	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size, 1,
 				 p->src_type, p->partition_vcpu_memory_access);
···
 		pipefds = malloc(sizeof(int) * nr_vcpus * 2);
 		TEST_ASSERT(pipefds, "Unable to allocate memory for pipefd");
 
-		for (vcpu_id = 0; vcpu_id < nr_vcpus; vcpu_id++) {
+		for (i = 0; i < nr_vcpus; i++) {
 			struct perf_test_vcpu_args *vcpu_args;
 			void *vcpu_hva;
 			void *vcpu_alias;
 
-			vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+			vcpu_args = &perf_test_args.vcpu_args[i];
 
 			/* Cache the host addresses of the region */
 			vcpu_hva = addr_gpa2hva(vm, vcpu_args->gpa);
···
 			 * Set up user fault fd to handle demand paging
 			 * requests.
 			 */
-			r = pipe2(&pipefds[vcpu_id * 2],
+			r = pipe2(&pipefds[i * 2],
 				  O_CLOEXEC | O_NONBLOCK);
 			TEST_ASSERT(!r, "Failed to set up pipefd");
 
-			setup_demand_paging(vm, &uffd_handler_threads[vcpu_id],
-					    pipefds[vcpu_id * 2], p->uffd_mode,
-					    p->uffd_delay, &uffd_args[vcpu_id],
+			setup_demand_paging(vm, &uffd_handler_threads[i],
+					    pipefds[i * 2], p->uffd_mode,
+					    p->uffd_delay, &uffd_args[i],
 					    vcpu_hva, vcpu_alias,
 					    vcpu_args->pages * perf_test_args.guest_page_size);
 		}
···
 		char c;
 
 		/* Tell the user fault fd handler threads to quit */
-		for (vcpu_id = 0; vcpu_id < nr_vcpus; vcpu_id++) {
-			r = write(pipefds[vcpu_id * 2 + 1], &c, 1);
+		for (i = 0; i < nr_vcpus; i++) {
+			r = write(pipefds[i * 2 + 1], &c, 1);
 			TEST_ASSERT(r == 1, "Unable to write to pipefd");
 
-			pthread_join(uffd_handler_threads[vcpu_id], NULL);
+			pthread_join(uffd_handler_threads[i], NULL);
 		}
 	}
 

tools/testing/selftests/kvm/dirty_log_perf_test.c

···
 
 static void vcpu_worker(struct perf_test_vcpu_args *vcpu_args)
 {
-	int ret;
+	struct kvm_vcpu *vcpu = vcpu_args->vcpu;
 	struct kvm_vm *vm = perf_test_args.vm;
+	int vcpu_idx = vcpu_args->vcpu_idx;
 	uint64_t pages_count = 0;
 	struct kvm_run *run;
 	struct timespec start;
 	struct timespec ts_diff;
 	struct timespec total = (struct timespec){0};
 	struct timespec avg;
-	int vcpu_id = vcpu_args->vcpu_id;
+	int ret;
 
-	run = vcpu_state(vm, vcpu_id);
+	run = vcpu->run;
 
 	while (!READ_ONCE(host_quit)) {
 		int current_iteration = READ_ONCE(iteration);
 
 		clock_gettime(CLOCK_MONOTONIC, &start);
-		ret = _vcpu_run(vm, vcpu_id);
+		ret = _vcpu_run(vm, vcpu->id);
 		ts_diff = timespec_elapsed(start);
 
 		TEST_ASSERT(ret == 0, "vcpu_run failed: %d\n", ret);
-		TEST_ASSERT(get_ucall(vm, vcpu_id, NULL) == UCALL_SYNC,
+		TEST_ASSERT(get_ucall(vm, vcpu->id, NULL) == UCALL_SYNC,
 			    "Invalid guest sync status: exit_reason=%s\n",
 			    exit_reason_str(run->exit_reason));
 
-		pr_debug("Got sync event from vCPU %d\n", vcpu_id);
-		vcpu_last_completed_iteration[vcpu_id] = current_iteration;
+		pr_debug("Got sync event from vCPU %d\n", vcpu_idx);
+		vcpu_last_completed_iteration[vcpu_idx] = current_iteration;
 		pr_debug("vCPU %d updated last completed iteration to %d\n",
-			 vcpu_id, vcpu_last_completed_iteration[vcpu_id]);
+			 vcpu->id, vcpu_last_completed_iteration[vcpu_idx]);
 
 		if (current_iteration) {
 			pages_count += vcpu_args->pages;
 			total = timespec_add(total, ts_diff);
 			pr_debug("vCPU %d iteration %d dirty memory time: %ld.%.9lds\n",
-				vcpu_id, current_iteration, ts_diff.tv_sec,
+				vcpu_idx, current_iteration, ts_diff.tv_sec,
 				ts_diff.tv_nsec);
 		} else {
 			pr_debug("vCPU %d iteration %d populate memory time: %ld.%.9lds\n",
-				vcpu_id, current_iteration, ts_diff.tv_sec,
+				vcpu_idx, current_iteration, ts_diff.tv_sec,
 				ts_diff.tv_nsec);
 		}
 
···
 		       !READ_ONCE(host_quit)) {}
 	}
 
-	avg = timespec_div(total, vcpu_last_completed_iteration[vcpu_id]);
+	avg = timespec_div(total, vcpu_last_completed_iteration[vcpu_idx]);
 	pr_debug("\nvCPU %d dirtied 0x%lx pages over %d iterations in %ld.%.9lds. (Avg %ld.%.9lds/iteration)\n",
-		vcpu_id, pages_count, vcpu_last_completed_iteration[vcpu_id],
+		vcpu_idx, pages_count, vcpu_last_completed_iteration[vcpu_idx],
 		total.tv_sec, total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 }
 
···
 	uint64_t guest_num_pages;
 	uint64_t host_num_pages;
 	uint64_t pages_per_slot;
-	int vcpu_id;
 	struct timespec start;
 	struct timespec ts_diff;
 	struct timespec get_dirty_log_total = (struct timespec){0};
 	struct timespec vcpu_dirty_total = (struct timespec){0};
 	struct timespec avg;
 	struct timespec clear_dirty_log_total = (struct timespec){0};
+	int i;
 
 	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size,
 				 p->slots, p->backing_src,
···
 	host_quit = false;
 
 	clock_gettime(CLOCK_MONOTONIC, &start);
-	for (vcpu_id = 0; vcpu_id < nr_vcpus; vcpu_id++)
-		vcpu_last_completed_iteration[vcpu_id] = -1;
+	for (i = 0; i < nr_vcpus; i++)
+		vcpu_last_completed_iteration[i] = -1;
 
 	perf_test_start_vcpu_threads(nr_vcpus, vcpu_worker);
 
 	/* Allow the vCPUs to populate memory */
 	pr_debug("Starting iteration %d - Populating\n", iteration);
-	for (vcpu_id = 0; vcpu_id < nr_vcpus; vcpu_id++) {
-		while (READ_ONCE(vcpu_last_completed_iteration[vcpu_id]) !=
+	for (i = 0; i < nr_vcpus; i++) {
+		while (READ_ONCE(vcpu_last_completed_iteration[i]) !=
 		       iteration)
 			;
 	}
···
 		iteration++;
 
 		pr_debug("Starting iteration %d\n", iteration);
-		for (vcpu_id = 0; vcpu_id < nr_vcpus; vcpu_id++) {
-			while (READ_ONCE(vcpu_last_completed_iteration[vcpu_id])
+		for (i = 0; i < nr_vcpus; i++) {
+			while (READ_ONCE(vcpu_last_completed_iteration[i])
 			       != iteration)
 				;
 		}

tools/testing/selftests/kvm/include/perf_test_util.h

···
 	uint64_t pages;
 
 	/* Only used by the host userspace part of the vCPU thread */
-	int vcpu_id;
+	struct kvm_vcpu *vcpu;
+	int vcpu_idx;
 };
 
 struct perf_test_args {
···
 
 extern struct perf_test_args perf_test_args;
 
-struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int nr_vcpus,
 				   uint64_t vcpu_memory_bytes, int slots,
 				   enum vm_mem_backing_src_type backing_src,
 				   bool partition_vcpu_memory_access);
···
 void perf_test_guest_code(uint32_t vcpu_id);
 
 uint64_t perf_test_nested_pages(int nr_vcpus);
-void perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus);
+void perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus, struct kvm_vcpu *vcpus[]);
 
 #endif /* SELFTEST_KVM_PERF_TEST_UTIL_H */

tools/testing/selftests/kvm/lib/perf_test_util.c

···
 static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
 
 struct vcpu_thread {
-	/* The id of the vCPU. */
-	int vcpu_id;
+	/* The index of the vCPU. */
+	int vcpu_idx;
 
 	/* The pthread backing the vCPU. */
 	pthread_t thread;
···
 /* Set to true once all vCPU threads are up and running. */
 static bool all_vcpu_threads_running;
 
+static struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
+
 /*
  * Continuously write to the first 8 bytes of each page in the
  * specified region.
  */
-void perf_test_guest_code(uint32_t vcpu_id)
+void perf_test_guest_code(uint32_t vcpu_idx)
 {
 	struct perf_test_args *pta = &perf_test_args;
-	struct perf_test_vcpu_args *vcpu_args = &pta->vcpu_args[vcpu_id];
+	struct perf_test_vcpu_args *vcpu_args = &pta->vcpu_args[vcpu_idx];
 	uint64_t gva;
 	uint64_t pages;
 	int i;
 
-	/* Make sure vCPU args data structure is not corrupt. */
-	GUEST_ASSERT(vcpu_args->vcpu_id == vcpu_id);
-
 	gva = vcpu_args->gva;
 	pages = vcpu_args->pages;
+
+	/* Make sure vCPU args data structure is not corrupt. */
+	GUEST_ASSERT(vcpu_args->vcpu_idx == vcpu_idx);
 
 	while (true) {
 		for (i = 0; i < pages; i++) {
···
 	}
 }
 
-void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus,
+void perf_test_setup_vcpus(struct kvm_vm *vm, int nr_vcpus,
+			   struct kvm_vcpu *vcpus[],
 			   uint64_t vcpu_memory_bytes,
 			   bool partition_vcpu_memory_access)
 {
 	struct perf_test_args *pta = &perf_test_args;
 	struct perf_test_vcpu_args *vcpu_args;
-	int vcpu_id;
+	int i;
 
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		vcpu_args = &pta->vcpu_args[vcpu_id];
+	for (i = 0; i < nr_vcpus; i++) {
+		vcpu_args = &pta->vcpu_args[i];
 
-		vcpu_args->vcpu_id = vcpu_id;
+		vcpu_args->vcpu = vcpus[i];
+		vcpu_args->vcpu_idx = i;
+
 		if (partition_vcpu_memory_access) {
 			vcpu_args->gva = guest_test_virt_mem +
-					 (vcpu_id * vcpu_memory_bytes);
+					 (i * vcpu_memory_bytes);
 			vcpu_args->pages = vcpu_memory_bytes /
 					   pta->guest_page_size;
-			vcpu_args->gpa = pta->gpa + (vcpu_id * vcpu_memory_bytes);
+			vcpu_args->gpa = pta->gpa + (i * vcpu_memory_bytes);
 		} else {
 			vcpu_args->gva = guest_test_virt_mem;
-			vcpu_args->pages = (vcpus * vcpu_memory_bytes) /
+			vcpu_args->pages = (nr_vcpus * vcpu_memory_bytes) /
 					   pta->guest_page_size;
 			vcpu_args->gpa = pta->gpa;
 		}
 
-		vcpu_args_set(vm, vcpu_id, 1, vcpu_id);
+		vcpu_args_set(vm, vcpus[i]->id, 1, i);
 
 		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
-			 vcpu_id, vcpu_args->gpa, vcpu_args->gpa +
+			 i, vcpu_args->gpa, vcpu_args->gpa +
 			 (vcpu_args->pages * pta->guest_page_size));
 	}
 }
 
-struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int nr_vcpus,
 				   uint64_t vcpu_memory_bytes, int slots,
 				   enum vm_mem_backing_src_type backing_src,
 				   bool partition_vcpu_memory_access)
···
 	pta->guest_page_size = vm_guest_mode_params[mode].page_size;
 
 	guest_num_pages = vm_adjust_num_guest_pages(mode,
-				(vcpus * vcpu_memory_bytes) / pta->guest_page_size);
+				(nr_vcpus * vcpu_memory_bytes) / pta->guest_page_size);
 
 	TEST_ASSERT(vcpu_memory_bytes % getpagesize() == 0,
 		    "Guest memory size is not host page size aligned.");
···
 	 * in-memory data structures.
 	 */
 	if (pta->nested)
-		slot0_pages += perf_test_nested_pages(vcpus);
+		slot0_pages += perf_test_nested_pages(nr_vcpus);
 
 	/*
 	 * Pass guest_num_pages to populate the page tables for test memory.
 	 * The memory is also added to memslot 0, but that's a benign side
 	 * effect as KVM allows aliasing HVAs in meslots.
 	 */
-	vm = __vm_create_with_vcpus(mode, vcpus, DEFAULT_GUEST_PHY_PAGES,
+	vm = __vm_create_with_vcpus(mode, nr_vcpus, DEFAULT_GUEST_PHY_PAGES,
 				    slot0_pages + guest_num_pages, 0,
-				    perf_test_guest_code, NULL);
+				    perf_test_guest_code, vcpus);
 
 	pta->vm = vm;
 
···
 	TEST_ASSERT(guest_num_pages < region_end_gfn,
 		    "Requested more guest memory than address space allows.\n"
 		    "    guest pages: %" PRIx64 " max gfn: %" PRIx64
-		    " vcpus: %d wss: %" PRIx64 "]\n",
-		    guest_num_pages, region_end_gfn - 1, vcpus,
-		    vcpu_memory_bytes);
+		    " nr_vcpus: %d wss: %" PRIx64 "]\n",
+		    guest_num_pages, region_end_gfn - 1, nr_vcpus, vcpu_memory_bytes);
 
-	pta->gpa = (region_end_gfn - guest_num_pages) * pta->guest_page_size;
+	pta->gpa = (region_end_gfn - guest_num_pages - 1) * pta->guest_page_size;
 	pta->gpa = align_down(pta->gpa, backing_src_pagesz);
 #ifdef __s390x__
 	/* Align to 1M (segment size) */
···
 	/* Do mapping for the demand paging memory slot */
 	virt_map(vm, guest_test_virt_mem, pta->gpa, guest_num_pages);
 
-	perf_test_setup_vcpus(vm, vcpus, vcpu_memory_bytes, partition_vcpu_memory_access);
+	perf_test_setup_vcpus(vm, nr_vcpus, vcpus, vcpu_memory_bytes,
+			      partition_vcpu_memory_access);
 
 	if (pta->nested) {
 		pr_info("Configuring vCPUs to run in L2 (nested).\n");
-		perf_test_setup_nested(vm, vcpus);
+		perf_test_setup_nested(vm, nr_vcpus, vcpus);
 	}
 
 	ucall_init(vm, NULL);
···
 	return 0;
 }
 
-void __weak perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus)
+void __weak perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus, struct kvm_vcpu **vcpus)
 {
 	pr_info("%s() not support on this architecture, skipping.\n", __func__);
 	exit(KSFT_SKIP);
···
 	while (!READ_ONCE(all_vcpu_threads_running))
 		;
 
-	vcpu_thread_fn(&perf_test_args.vcpu_args[vcpu->vcpu_id]);
+	vcpu_thread_fn(&perf_test_args.vcpu_args[vcpu->vcpu_idx]);
 
 	return NULL;
 }
 
-void perf_test_start_vcpu_threads(int vcpus, void (*vcpu_fn)(struct perf_test_vcpu_args *))
+void perf_test_start_vcpu_threads(int nr_vcpus,
+				  void (*vcpu_fn)(struct perf_test_vcpu_args *))
 {
-	int vcpu_id;
+	int i;
 
 	vcpu_thread_fn = vcpu_fn;
 	WRITE_ONCE(all_vcpu_threads_running, false);
 
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		struct vcpu_thread *vcpu = &vcpu_threads[vcpu_id];
+	for (i = 0; i < nr_vcpus; i++) {
+		struct vcpu_thread *vcpu = &vcpu_threads[i];
 
-		vcpu->vcpu_id = vcpu_id;
+		vcpu->vcpu_idx = i;
 		WRITE_ONCE(vcpu->running, false);
 
 		pthread_create(&vcpu->thread, NULL, vcpu_thread_main, vcpu);
 	}
 
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		while (!READ_ONCE(vcpu_threads[vcpu_id].running))
+	for (i = 0; i < nr_vcpus; i++) {
+		while (!READ_ONCE(vcpu_threads[i].running))
 			;
 	}
 
 	WRITE_ONCE(all_vcpu_threads_running, true);
 }
 
-void perf_test_join_vcpu_threads(int vcpus)
+void perf_test_join_vcpu_threads(int nr_vcpus)
 {
-	int vcpu_id;
+	int i;
 
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++)
-		pthread_join(vcpu_threads[vcpu_id].thread, NULL);
+	for (i = 0; i < nr_vcpus; i++)
+		pthread_join(vcpu_threads[i].thread, NULL);
 }

tools/testing/selftests/kvm/lib/x86_64/perf_test_util.c

···
 	nested_identity_map_1g(vmx, vm, start, end - start);
 }
 
-void perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus)
+void perf_test_setup_nested(struct kvm_vm *vm, int nr_vcpus, struct kvm_vcpu *vcpus[])
 {
 	struct vmx_pages *vmx, *vmx0 = NULL;
 	struct kvm_regs regs;
···
 		 * Override the vCPU to run perf_test_l1_guest_code() which will
 		 * bounce it into L2 before calling perf_test_guest_code().
 		 */
-		vcpu_regs_get(vm, vcpu_id, &regs);
+		vcpu_regs_get(vm, vcpus[vcpu_id]->id, &regs);
 		regs.rip = (unsigned long) perf_test_l1_guest_code;
-		vcpu_regs_set(vm, vcpu_id, &regs);
-		vcpu_args_set(vm, vcpu_id, 2, vmx_gva, vcpu_id);
+		vcpu_regs_set(vm, vcpus[vcpu_id]->id, &regs);
+		vcpu_args_set(vm, vcpus[vcpu_id]->id, 2, vmx_gva, vcpu_id);
 	}
 }

tools/testing/selftests/kvm/memslot_modification_stress_test.c

···
 
 static void vcpu_worker(struct perf_test_vcpu_args *vcpu_args)
 {
-	int ret;
-	int vcpu_id = vcpu_args->vcpu_id;
+	struct kvm_vcpu *vcpu = vcpu_args->vcpu;
 	struct kvm_vm *vm = perf_test_args.vm;
 	struct kvm_run *run;
+	int ret;
 
-	run = vcpu_state(vm, vcpu_id);
+	run = vcpu->run;
 
 	/* Let the guest access its memory until a stop signal is received */
 	while (READ_ONCE(run_vcpus)) {
-		ret = _vcpu_run(vm, vcpu_id);
+		ret = _vcpu_run(vm, vcpu->id);
 		TEST_ASSERT(ret == 0, "vcpu_run failed: %d\n", ret);
 
-		if (get_ucall(vm, vcpu_id, NULL) == UCALL_SYNC)
+		if (get_ucall(vm, vcpu->id, NULL) == UCALL_SYNC)
 			continue;
 
 		TEST_ASSERT(false,