1 files changed, 269 insertions, 0 deletions
diff --git a/tools/testing/selftests/kvm/x86_64/nx_huge_pages_test.c b/tools/testing/selftests/kvm/x86_64/nx_huge_pages_test.c
new file mode 100644
index 000000000000..17bbb96fc4df
--- /dev/null
+++ b/tools/testing/selftests/kvm/x86_64/nx_huge_pages_test.c
@@ -0,0 +1,269 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Usage: to be run via nx_huge_page_test.sh, which does the necessary
+ * environment setup and teardown
+ *
+ * Copyright (C) 2022, Google LLC.
+ */
+
+#define _GNU_SOURCE
+
+#include <fcntl.h>
+#include <stdint.h>
+#include <time.h>
+
+#include <test_util.h>
+#include "kvm_util.h"
+#include "processor.h"
+
+#define HPAGE_SLOT		10
+#define HPAGE_GPA		(4UL << 30) /* 4G prevents collision w/ slot 0 */
+#define HPAGE_GVA		HPAGE_GPA /* GVA is arbitrary, so use GPA. */
+#define PAGES_PER_2MB_HUGE_PAGE 512
+#define HPAGE_SLOT_NPAGES	(3 * PAGES_PER_2MB_HUGE_PAGE)
+
+/*
+ * Passed by nx_huge_pages_test.sh to provide an easy warning if this test is
+ * being run without it.
+ */
+#define MAGIC_TOKEN 887563923
+
+/*
+ * x86 opcode for the return instruction. Used to call into, and then
+ * immediately return from, memory backed with hugepages.
+ */
+#define RETURN_OPCODE 0xC3
+
+/* Call the specified memory address. */
+static void guest_do_CALL(uint64_t target)
+{
+	((void (*)(void)) target)();
+}
+
+/*
+ * Exit the VM after each memory access so that the userspace component of the
+ * test can make assertions about the pages backing the VM.
+ *
+ * See the below for an explanation of how each access should affect the
+ * backing mappings.
+ */
+void guest_code(void)
+{
+	uint64_t hpage_1 = HPAGE_GVA;
+	uint64_t hpage_2 = hpage_1 + (PAGE_SIZE * 512);
+	uint64_t hpage_3 = hpage_2 + (PAGE_SIZE * 512);
+
+	READ_ONCE(*(uint64_t *)hpage_1);
+	GUEST_SYNC(1);
+
+	READ_ONCE(*(uint64_t *)hpage_2);
+	GUEST_SYNC(2);
+
+	guest_do_CALL(hpage_1);
+	GUEST_SYNC(3);
+
+	guest_do_CALL(hpage_3);
+	GUEST_SYNC(4);
+
+	READ_ONCE(*(uint64_t *)hpage_1);
+	GUEST_SYNC(5);
+
+	READ_ONCE(*(uint64_t *)hpage_3);
+	GUEST_SYNC(6);
+}
+
+static void check_2m_page_count(struct kvm_vm *vm, int expected_pages_2m)
+{
+	int actual_pages_2m;
+
+	actual_pages_2m = vm_get_stat(vm, "pages_2m");
+
+	TEST_ASSERT(actual_pages_2m == expected_pages_2m,
+		    "Unexpected 2m page count. Expected %d, got %d",
+		    expected_pages_2m, actual_pages_2m);
+}
+
+static void check_split_count(struct kvm_vm *vm, int expected_splits)
+{
+	int actual_splits;
+
+	actual_splits = vm_get_stat(vm, "nx_lpage_splits");
+
+	TEST_ASSERT(actual_splits == expected_splits,
+		    "Unexpected NX huge page split count. Expected %d, got %d",
+		    expected_splits, actual_splits);
+}
+
+static void wait_for_reclaim(int reclaim_period_ms)
+{
+	long reclaim_wait_ms;
+	struct timespec ts;
+
+	reclaim_wait_ms = reclaim_period_ms * 5;
+	ts.tv_sec = reclaim_wait_ms / 1000;
+	ts.tv_nsec = (reclaim_wait_ms - (ts.tv_sec * 1000)) * 1000000;
+	nanosleep(&ts, NULL);
+}
+
+void run_test(int reclaim_period_ms, bool disable_nx_huge_pages,
+	      bool reboot_permissions)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	uint64_t nr_bytes;
+	void *hva;
+	int r;
+
+	vm = vm_create(1);
+
+	if (disable_nx_huge_pages) {
+		r = __vm_disable_nx_huge_pages(vm);
+		if (reboot_permissions) {
+			TEST_ASSERT(!r, "Disabling NX huge pages should succeed if process has reboot permissions");
+		} else {
+			TEST_ASSERT(r == -1 && errno == EPERM,
+				    "This process should not have permission to disable NX huge pages");
+			return;
+		}
+	}
+
+	vcpu = vm_vcpu_add(vm, 0, guest_code);
+
+	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS_HUGETLB,
+				    HPAGE_GPA, HPAGE_SLOT,
+				    HPAGE_SLOT_NPAGES, 0);
+
+	nr_bytes = HPAGE_SLOT_NPAGES * vm->page_size;
+
+	/*
+	 * Ensure that KVM can map HPAGE_SLOT with huge pages by mapping the
+	 * region into the guest with 2MiB pages whenever TDP is disabled (i.e.
+	 * whenever KVM is shadowing the guest page tables).
+	 *
+	 * When TDP is enabled, KVM should be able to map HPAGE_SLOT with huge
+	 * pages irrespective of the guest page size, so map with 4KiB pages
+	 * to test that that is the case.
+	 */
+	if (kvm_is_tdp_enabled())
+		virt_map_level(vm, HPAGE_GVA, HPAGE_GPA, nr_bytes, PG_LEVEL_4K);
+	else
+		virt_map_level(vm, HPAGE_GVA, HPAGE_GPA, nr_bytes, PG_LEVEL_2M);
+
+	hva = addr_gpa2hva(vm, HPAGE_GPA);
+	memset(hva, RETURN_OPCODE, nr_bytes);
+
+	check_2m_page_count(vm, 0);
+	check_split_count(vm, 0);
+
+	/*
+	 * The guest code will first read from the first hugepage, resulting
+	 * in a huge page mapping being created.
+	 */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, 1);
+	check_split_count(vm, 0);
+
+	/*
+	 * Then the guest code will read from the second hugepage, resulting
+	 * in another huge page mapping being created.
+	 */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, 2);
+	check_split_count(vm, 0);
+
+	/*
+	 * Next, the guest will execute from the first huge page, causing it
+	 * to be remapped at 4k.
+	 *
+	 * If NX huge pages are disabled, this should have no effect.
+	 */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, disable_nx_huge_pages ? 2 : 1);
+	check_split_count(vm, disable_nx_huge_pages ? 0 : 1);
+
+	/*
+	 * Executing from the third huge page (previously unaccessed) will
+	 * cause part to be mapped at 4k.
+	 *
+	 * If NX huge pages are disabled, it should be mapped at 2M.
+	 */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, disable_nx_huge_pages ? 3 : 1);
+	check_split_count(vm, disable_nx_huge_pages ? 0 : 2);
+
+	/* Reading from the first huge page again should have no effect. */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, disable_nx_huge_pages ? 3 : 1);
+	check_split_count(vm, disable_nx_huge_pages ? 0 : 2);
+
+	/* Give recovery thread time to run. */
+	wait_for_reclaim(reclaim_period_ms);
+
+	/*
+	 * Now that the reclaimer has run, all the split pages should be gone.
+	 *
+	 * If NX huge pages are disabled, the relaimer will not run, so
+	 * nothing should change from here on.
+	 */
+	check_2m_page_count(vm, disable_nx_huge_pages ? 3 : 1);
+	check_split_count(vm, 0);
+
+	/*
+	 * The 4k mapping on hpage 3 should have been removed, so check that
+	 * reading from it causes a huge page mapping to be installed.
+	 */
+	vcpu_run(vcpu);
+	check_2m_page_count(vm, disable_nx_huge_pages ? 3 : 2);
+	check_split_count(vm, 0);
+
+	kvm_vm_free(vm);
+}
+
+static void help(char *name)
+{
+	puts("");
+	printf("usage: %s [-h] [-p period_ms] [-t token]\n", name);
+	puts("");
+	printf(" -p: The NX reclaim period in milliseconds.\n");
+	printf(" -t: The magic token to indicate environment setup is done.\n");
+	printf(" -r: The test has reboot permissions and can disable NX huge pages.\n");
+	puts("");
+	exit(0);
+}
+
+int main(int argc, char **argv)
+{
+	int reclaim_period_ms = 0, token = 0, opt;
+	bool reboot_permissions = false;
+
+	while ((opt = getopt(argc, argv, "hp:t:r")) != -1) {
+		switch (opt) {
+		case 'p':
+			reclaim_period_ms = atoi_positive("Reclaim period", optarg);
+			break;
+		case 't':
+			token = atoi_paranoid(optarg);
+			break;
+		case 'r':
+			reboot_permissions = true;
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			break;
+		}
+	}
+
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_VM_DISABLE_NX_HUGE_PAGES));
+
+	__TEST_REQUIRE(token == MAGIC_TOKEN,
+		       "This test must be run with the magic token via '-t %d'.\n"
+		       "Running via nx_huge_pages_test.sh, which also handles "
+		       "environment setup, is strongly recommended.", MAGIC_TOKEN);
+
+	run_test(reclaim_period_ms, false, reboot_permissions);
+	run_test(reclaim_period_ms, true, reboot_permissions);
+
+	return 0;
+}
+