4 files changed, 168 insertions, 68 deletions

diff --git a/tools/testing/selftests/kvm/lib/elf.c b/tools/testing/selftests/kvm/lib/elf.c
index eac44f5d0db0..13e8e3dcf984 100644
--- a/tools/testing/selftests/kvm/lib/elf.c
+++ b/tools/testing/selftests/kvm/lib/elf.c
@@ -157,8 +157,7 @@ void kvm_vm_elf_load(struct kvm_vm *vm, const char *filename)
 			"memsize of 0,\n"
 			"  phdr index: %u p_memsz: 0x%" PRIx64,
 			n1, (uint64_t) phdr.p_memsz);
-		vm_vaddr_t seg_vstart = phdr.p_vaddr;
-		seg_vstart &= ~(vm_vaddr_t)(vm->page_size - 1);
+		vm_vaddr_t seg_vstart = align_down(phdr.p_vaddr, vm->page_size);
 		vm_vaddr_t seg_vend = phdr.p_vaddr + phdr.p_memsz - 1;
 		seg_vend |= vm->page_size - 1;
 		size_t seg_size = seg_vend - seg_vstart + 1;
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index 14bb4d5b6bb7..8f2e0bb1ef96 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -22,15 +22,6 @@
 
 static int vcpu_mmap_sz(void);
 
-/* Aligns x up to the next multiple of size. Size must be a power of 2. */
-static void *align(void *x, size_t size)
-{
-	size_t mask = size - 1;
-	TEST_ASSERT(size != 0 && !(size & (size - 1)),
-		    "size not a power of 2: %lu", size);
-	return (void *) (((size_t) x + mask) & ~mask);
-}
-
 int open_path_or_exit(const char *path, int flags)
 {
 	int fd;
@@ -191,15 +182,15 @@ const char *vm_guest_mode_string(uint32_t i)
 }
 
 const struct vm_guest_mode_params vm_guest_mode_params[] = {
-	{ 52, 48,  0x1000, 12 },
-	{ 52, 48, 0x10000, 16 },
-	{ 48, 48,  0x1000, 12 },
-	{ 48, 48, 0x10000, 16 },
-	{ 40, 48,  0x1000, 12 },
-	{ 40, 48, 0x10000, 16 },
-	{  0,  0,  0x1000, 12 },
-	{ 47, 64,  0x1000, 12 },
-	{ 44, 64,  0x1000, 12 },
+	[VM_MODE_P52V48_4K]	= { 52, 48,  0x1000, 12 },
+	[VM_MODE_P52V48_64K]	= { 52, 48, 0x10000, 16 },
+	[VM_MODE_P48V48_4K]	= { 48, 48,  0x1000, 12 },
+	[VM_MODE_P48V48_64K]	= { 48, 48, 0x10000, 16 },
+	[VM_MODE_P40V48_4K]	= { 40, 48,  0x1000, 12 },
+	[VM_MODE_P40V48_64K]	= { 40, 48, 0x10000, 16 },
+	[VM_MODE_PXXV48_4K]	= {  0,  0,  0x1000, 12 },
+	[VM_MODE_P47V64_4K]	= { 47, 64,  0x1000, 12 },
+	[VM_MODE_P44V64_4K]	= { 44, 64,  0x1000, 12 },
 };
 _Static_assert(sizeof(vm_guest_mode_params)/sizeof(struct vm_guest_mode_params) == NUM_VM_MODES,
 	       "Missing new mode params?");
@@ -879,9 +870,17 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 	alignment = 1;
 #endif
 
+	/*
+	 * When using THP mmap is not guaranteed to returned a hugepage aligned
+	 * address so we have to pad the mmap. Padding is not needed for HugeTLB
+	 * because mmap will always return an address aligned to the HugeTLB
+	 * page size.
+	 */
 	if (src_type == VM_MEM_SRC_ANONYMOUS_THP)
 		alignment = max(backing_src_pagesz, alignment);
 
+	ASSERT_EQ(guest_paddr, align_up(guest_paddr, backing_src_pagesz));
+
 	/* Add enough memory to align up if necessary */
 	if (alignment > 1)
 		region->mmap_size += alignment;
@@ -914,8 +913,13 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 		    "test_malloc failed, mmap_start: %p errno: %i",
 		    region->mmap_start, errno);
 
+	TEST_ASSERT(!is_backing_src_hugetlb(src_type) ||
+		    region->mmap_start == align_ptr_up(region->mmap_start, backing_src_pagesz),
+		    "mmap_start %p is not aligned to HugeTLB page size 0x%lx",
+		    region->mmap_start, backing_src_pagesz);
+
 	/* Align host address */
-	region->host_mem = align(region->mmap_start, alignment);
+	region->host_mem = align_ptr_up(region->mmap_start, alignment);
 
 	/* As needed perform madvise */
 	if ((src_type == VM_MEM_SRC_ANONYMOUS ||
@@ -958,7 +962,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 			    "mmap of alias failed, errno: %i", errno);
 
 		/* Align host alias address */
-		region->host_alias = align(region->mmap_alias, alignment);
+		region->host_alias = align_ptr_up(region->mmap_alias, alignment);
 	}
 }
 
diff --git a/tools/testing/selftests/kvm/lib/perf_test_util.c b/tools/testing/selftests/kvm/lib/perf_test_util.c
index 0ef80dbdc116..722df3a28791 100644
--- a/tools/testing/selftests/kvm/lib/perf_test_util.c
+++ b/tools/testing/selftests/kvm/lib/perf_test_util.c
@@ -10,21 +10,40 @@
 
 struct perf_test_args perf_test_args;
 
-uint64_t guest_test_phys_mem;
-
 /*
  * Guest virtual memory offset of the testing memory slot.
  * Must not conflict with identity mapped test code.
  */
 static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
 
+struct vcpu_thread {
+	/* The id of the vCPU. */
+	int vcpu_id;
+
+	/* The pthread backing the vCPU. */
+	pthread_t thread;
+
+	/* Set to true once the vCPU thread is up and running. */
+	bool running;
+};
+
+/* The vCPU threads involved in this test. */
+static struct vcpu_thread vcpu_threads[KVM_MAX_VCPUS];
+
+/* The function run by each vCPU thread, as provided by the test. */
+static void (*vcpu_thread_fn)(struct perf_test_vcpu_args *);
+
+/* Set to true once all vCPU threads are up and running. */
+static bool all_vcpu_threads_running;
+
 /*
  * Continuously write to the first 8 bytes of each page in the
  * specified region.
  */
 static void guest_code(uint32_t vcpu_id)
 {
-	struct perf_test_vcpu_args *vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+	struct perf_test_args *pta = &perf_test_args;
+	struct perf_test_vcpu_args *vcpu_args = &pta->vcpu_args[vcpu_id];
 	uint64_t gva;
 	uint64_t pages;
 	int i;
@@ -37,9 +56,9 @@ static void guest_code(uint32_t vcpu_id)
 
 	while (true) {
 		for (i = 0; i < pages; i++) {
-			uint64_t addr = gva + (i * perf_test_args.guest_page_size);
+			uint64_t addr = gva + (i * pta->guest_page_size);
 
-			if (i % perf_test_args.wr_fract == 0)
+			if (i % pta->wr_fract == 0)
 				*(uint64_t *)addr = 0x0123456789ABCDEF;
 			else
 				READ_ONCE(*(uint64_t *)addr);
@@ -49,35 +68,81 @@ static void guest_code(uint32_t vcpu_id)
 	}
 }
 
+void perf_test_setup_vcpus(struct kvm_vm *vm, int 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;
+
+	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
+		vcpu_args = &pta->vcpu_args[vcpu_id];
+
+		vcpu_args->vcpu_id = vcpu_id;
+		if (partition_vcpu_memory_access) {
+			vcpu_args->gva = guest_test_virt_mem +
+					 (vcpu_id * vcpu_memory_bytes);
+			vcpu_args->pages = vcpu_memory_bytes /
+					   pta->guest_page_size;
+			vcpu_args->gpa = pta->gpa + (vcpu_id * vcpu_memory_bytes);
+		} else {
+			vcpu_args->gva = guest_test_virt_mem;
+			vcpu_args->pages = (vcpus * vcpu_memory_bytes) /
+					   pta->guest_page_size;
+			vcpu_args->gpa = pta->gpa;
+		}
+
+		vcpu_args_set(vm, vcpu_id, 1, vcpu_id);
+
+		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
+			 vcpu_id, 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,
 				   uint64_t vcpu_memory_bytes, int slots,
-				   enum vm_mem_backing_src_type backing_src)
+				   enum vm_mem_backing_src_type backing_src,
+				   bool partition_vcpu_memory_access)
 {
+	struct perf_test_args *pta = &perf_test_args;
 	struct kvm_vm *vm;
 	uint64_t guest_num_pages;
+	uint64_t backing_src_pagesz = get_backing_src_pagesz(backing_src);
 	int i;
 
 	pr_info("Testing guest mode: %s\n", vm_guest_mode_string(mode));
 
-	perf_test_args.host_page_size = getpagesize();
-	perf_test_args.guest_page_size = vm_guest_mode_params[mode].page_size;
+	/* By default vCPUs will write to memory. */
+	pta->wr_fract = 1;
+
+	/*
+	 * Snapshot the non-huge page size.  This is used by the guest code to
+	 * access/dirty pages at the logging granularity.
+	 */
+	pta->guest_page_size = vm_guest_mode_params[mode].page_size;
 
 	guest_num_pages = vm_adjust_num_guest_pages(mode,
-				(vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size);
+				(vcpus * vcpu_memory_bytes) / pta->guest_page_size);
 
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.host_page_size == 0,
+	TEST_ASSERT(vcpu_memory_bytes % getpagesize() == 0,
 		    "Guest memory size is not host page size aligned.");
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.guest_page_size == 0,
+	TEST_ASSERT(vcpu_memory_bytes % pta->guest_page_size == 0,
 		    "Guest memory size is not guest page size aligned.");
 	TEST_ASSERT(guest_num_pages % slots == 0,
 		    "Guest memory cannot be evenly divided into %d slots.",
 		    slots);
 
+	/*
+	 * 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,
-				  (vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size,
-				  0, guest_code, NULL);
+				  guest_num_pages, 0, guest_code, NULL);
 
-	perf_test_args.vm = vm;
+	pta->vm = vm;
 
 	/*
 	 * If there should be more memory in the guest test region than there
@@ -90,20 +155,18 @@ struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
 		    guest_num_pages, vm_get_max_gfn(vm), vcpus,
 		    vcpu_memory_bytes);
 
-	guest_test_phys_mem = (vm_get_max_gfn(vm) - guest_num_pages) *
-			      perf_test_args.guest_page_size;
-	guest_test_phys_mem &= ~(perf_test_args.host_page_size - 1);
+	pta->gpa = (vm_get_max_gfn(vm) - guest_num_pages) * pta->guest_page_size;
+	pta->gpa = align_down(pta->gpa, backing_src_pagesz);
 #ifdef __s390x__
 	/* Align to 1M (segment size) */
-	guest_test_phys_mem &= ~((1 << 20) - 1);
+	pta->gpa = align_down(pta->gpa, 1 << 20);
 #endif
-	pr_info("guest physical test memory offset: 0x%lx\n", guest_test_phys_mem);
+	pr_info("guest physical test memory offset: 0x%lx\n", pta->gpa);
 
 	/* Add extra memory slots for testing */
 	for (i = 0; i < slots; i++) {
 		uint64_t region_pages = guest_num_pages / slots;
-		vm_paddr_t region_start = guest_test_phys_mem +
-			region_pages * perf_test_args.guest_page_size * i;
+		vm_paddr_t region_start = pta->gpa + region_pages * pta->guest_page_size * i;
 
 		vm_userspace_mem_region_add(vm, backing_src, region_start,
 					    PERF_TEST_MEM_SLOT_INDEX + i,
@@ -111,10 +174,15 @@ struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
 	}
 
 	/* Do mapping for the demand paging memory slot */
-	virt_map(vm, guest_test_virt_mem, guest_test_phys_mem, guest_num_pages);
+	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);
 
 	ucall_init(vm, NULL);
 
+	/* Export the shared variables to the guest. */
+	sync_global_to_guest(vm, perf_test_args);
+
 	return vm;
 }
 
@@ -124,36 +192,60 @@ void perf_test_destroy_vm(struct kvm_vm *vm)
 	kvm_vm_free(vm);
 }
 
-void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus,
-			   uint64_t vcpu_memory_bytes,
-			   bool partition_vcpu_memory_access)
+void perf_test_set_wr_fract(struct kvm_vm *vm, int wr_fract)
+{
+	perf_test_args.wr_fract = wr_fract;
+	sync_global_to_guest(vm, perf_test_args);
+}
+
+static void *vcpu_thread_main(void *data)
+{
+	struct vcpu_thread *vcpu = data;
+
+	WRITE_ONCE(vcpu->running, true);
+
+	/*
+	 * Wait for all vCPU threads to be up and running before calling the test-
+	 * provided vCPU thread function. This prevents thread creation (which
+	 * requires taking the mmap_sem in write mode) from interfering with the
+	 * guest faulting in its memory.
+	 */
+	while (!READ_ONCE(all_vcpu_threads_running))
+		;
+
+	vcpu_thread_fn(&perf_test_args.vcpu_args[vcpu->vcpu_id]);
+
+	return NULL;
+}
+
+void perf_test_start_vcpu_threads(int vcpus, void (*vcpu_fn)(struct perf_test_vcpu_args *))
 {
-	vm_paddr_t vcpu_gpa;
-	struct perf_test_vcpu_args *vcpu_args;
 	int vcpu_id;
 
+	vcpu_thread_fn = vcpu_fn;
+	WRITE_ONCE(all_vcpu_threads_running, false);
+
 	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+		struct vcpu_thread *vcpu = &vcpu_threads[vcpu_id];
 
-		vcpu_args->vcpu_id = vcpu_id;
-		if (partition_vcpu_memory_access) {
-			vcpu_args->gva = guest_test_virt_mem +
-					 (vcpu_id * vcpu_memory_bytes);
-			vcpu_args->pages = vcpu_memory_bytes /
-					   perf_test_args.guest_page_size;
-			vcpu_gpa = guest_test_phys_mem +
-				   (vcpu_id * vcpu_memory_bytes);
-		} else {
-			vcpu_args->gva = guest_test_virt_mem;
-			vcpu_args->pages = (vcpus * vcpu_memory_bytes) /
-					   perf_test_args.guest_page_size;
-			vcpu_gpa = guest_test_phys_mem;
-		}
+		vcpu->vcpu_id = vcpu_id;
+		WRITE_ONCE(vcpu->running, false);
 
-		vcpu_args_set(vm, vcpu_id, 1, vcpu_id);
+		pthread_create(&vcpu->thread, NULL, vcpu_thread_main, vcpu);
+	}
 
-		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
-			 vcpu_id, vcpu_gpa, vcpu_gpa +
-			 (vcpu_args->pages * perf_test_args.guest_page_size));
+	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
+		while (!READ_ONCE(vcpu_threads[vcpu_id].running))
+			;
 	}
+
+	WRITE_ONCE(all_vcpu_threads_running, true);
+}
+
+void perf_test_join_vcpu_threads(int vcpus)
+{
+	int vcpu_id;
+
+	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++)
+		pthread_join(vcpu_threads[vcpu_id].thread, NULL);
 }
diff --git a/tools/testing/selftests/kvm/lib/test_util.c b/tools/testing/selftests/kvm/lib/test_util.c
index b72429108993..6d23878bbfe1 100644
--- a/tools/testing/selftests/kvm/lib/test_util.c
+++ b/tools/testing/selftests/kvm/lib/test_util.c
@@ -283,6 +283,11 @@ size_t get_backing_src_pagesz(uint32_t i)
 	}
 }
 
+bool is_backing_src_hugetlb(uint32_t i)
+{
+	return !!(vm_mem_backing_src_alias(i)->flag & MAP_HUGETLB);
+}
+
 static void print_available_backing_src_types(const char *prefix)
 {
 	int i;