KVM: selftests: add Enlightened VMCS test

Modify test library and add eVMCS test. This includes nVMX save/restore
testing.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

tools/testing/selftests/kvm/Makefile

@@ -13,6 +13,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/sync_regs_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_tsc_adjust_test
 TEST_GEN_PROGS_x86_64 += x86_64/cr4_cpuid_sync_test
 TEST_GEN_PROGS_x86_64 += x86_64/state_test
+TEST_GEN_PROGS_x86_64 += x86_64/evmcs_test
 TEST_GEN_PROGS_x86_64 += dirty_log_test
 
 TEST_GEN_PROGS_aarch64 += dirty_log_test

tools/testing/selftests/kvm/include/x86_64/vmx.h

@@ -339,6 +339,8 @@ struct vmx_msr_entry {
 	uint64_t value;
 } __attribute__ ((aligned(16)));
 
+#include "evmcs.h"
+
 static inline int vmxon(uint64_t phys)
 {
 	uint8_t ret;
@@ -372,6 +374,9 @@ static inline int vmptrld(uint64_t vmcs_pa)
 {
 	uint8_t ret;
 
+	if (enable_evmcs)
+		return -1;
+
 	__asm__ __volatile__ ("vmptrld %[pa]; setna %[ret]"
 		: [ret]"=rm"(ret)
 		: [pa]"m"(vmcs_pa)
@@ -385,6 +390,9 @@ static inline int vmptrst(uint64_t *value)
 	uint64_t tmp;
 	uint8_t ret;
 
+	if (enable_evmcs)
+		return evmcs_vmptrst(value);
+
 	__asm__ __volatile__("vmptrst %[value]; setna %[ret]"
 		: [value]"=m"(tmp), [ret]"=rm"(ret)
 		: : "cc", "memory");
@@ -411,6 +419,9 @@ static inline int vmlaunch(void)
 {
 	int ret;
 
+	if (enable_evmcs)
+		return evmcs_vmlaunch();
+
 	__asm__ __volatile__("push %%rbp;"
 			     "push %%rcx;"
 			     "push %%rdx;"
@@ -443,6 +454,9 @@ static inline int vmresume(void)
 {
 	int ret;
 
+	if (enable_evmcs)
+		return evmcs_vmresume();
+
 	__asm__ __volatile__("push %%rbp;"
 			     "push %%rcx;"
 			     "push %%rdx;"
@@ -482,6 +496,9 @@ static inline int vmread(uint64_t encoding, uint64_t *value)
 	uint64_t tmp;
 	uint8_t ret;
 
+	if (enable_evmcs)
+		return evmcs_vmread(encoding, value);
+
 	__asm__ __volatile__("vmread %[encoding], %[value]; setna %[ret]"
 		: [value]"=rm"(tmp), [ret]"=rm"(ret)
 		: [encoding]"r"(encoding)
@@ -506,6 +523,9 @@ static inline int vmwrite(uint64_t encoding, uint64_t value)
 {
 	uint8_t ret;
 
+	if (enable_evmcs)
+		return evmcs_vmwrite(encoding, value);
+
 	__asm__ __volatile__ ("vmwrite %[value], %[encoding]; setna %[ret]"
 		: [ret]"=rm"(ret)
 		: [value]"rm"(value), [encoding]"r"(encoding)
@@ -543,6 +563,14 @@ struct vmx_pages {
 	void *vmwrite_hva;
 	uint64_t vmwrite_gpa;
 	void *vmwrite;
+
+	void *vp_assist_hva;
+	uint64_t vp_assist_gpa;
+	void *vp_assist;
+
+	void *enlightened_vmcs_hva;
+	uint64_t enlightened_vmcs_gpa;
+	void *enlightened_vmcs;
 };
 
 struct vmx_pages *vcpu_alloc_vmx(struct kvm_vm *vm, vm_vaddr_t *p_vmx_gva);

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

@@ -13,6 +13,8 @@
 #include "processor.h"
 #include "vmx.h"
 
+bool enable_evmcs;
+
 /* Allocate memory regions for nested VMX tests.
  *
  * Input Args:
@@ -62,6 +64,20 @@ vcpu_alloc_vmx(struct kvm_vm *vm, vm_vaddr_t *p_vmx_gva)
 	vmx->vmwrite_gpa = addr_gva2gpa(vm, (uintptr_t)vmx->vmwrite);
 	memset(vmx->vmwrite_hva, 0, getpagesize());
 
+	/* Setup of a region of guest memory for the VP Assist page. */
+	vmx->vp_assist = (void *)vm_vaddr_alloc(vm, getpagesize(),
+						0x10000, 0, 0);
+	vmx->vp_assist_hva = addr_gva2hva(vm, (uintptr_t)vmx->vp_assist);
+	vmx->vp_assist_gpa = addr_gva2gpa(vm, (uintptr_t)vmx->vp_assist);
+
+	/* Setup of a region of guest memory for the enlightened VMCS. */
+	vmx->enlightened_vmcs = (void *)vm_vaddr_alloc(vm, getpagesize(),
+						       0x10000, 0, 0);
+	vmx->enlightened_vmcs_hva =
+		addr_gva2hva(vm, (uintptr_t)vmx->enlightened_vmcs);
+	vmx->enlightened_vmcs_gpa =
+		addr_gva2gpa(vm, (uintptr_t)vmx->enlightened_vmcs);
+
 	*p_vmx_gva = vmx_gva;
 	return vmx;
 }
@@ -112,18 +128,26 @@ bool prepare_for_vmx_operation(struct vmx_pages *vmx)
 
 bool load_vmcs(struct vmx_pages *vmx)
 {
-	/* Load a VMCS. */
-	*(uint32_t *)(vmx->vmcs) = vmcs_revision();
-	if (vmclear(vmx->vmcs_gpa))
-		return false;
-
-	if (vmptrld(vmx->vmcs_gpa))
-		return false;
-
-	/* Setup shadow VMCS, do not load it yet. */
-	*(uint32_t *)(vmx->shadow_vmcs) = vmcs_revision() | 0x80000000ul;
-	if (vmclear(vmx->shadow_vmcs_gpa))
-		return false;
+	if (!enable_evmcs) {
+		/* Load a VMCS. */
+		*(uint32_t *)(vmx->vmcs) = vmcs_revision();
+		if (vmclear(vmx->vmcs_gpa))
+			return false;
+
+		if (vmptrld(vmx->vmcs_gpa))
+			return false;
+
+		/* Setup shadow VMCS, do not load it yet. */
+		*(uint32_t *)(vmx->shadow_vmcs) =
+			vmcs_revision() | 0x80000000ul;
+		if (vmclear(vmx->shadow_vmcs_gpa))
+			return false;
+	} else {
+		if (evmcs_vmptrld(vmx->enlightened_vmcs_gpa,
+				  vmx->enlightened_vmcs))
+			return false;
+		current_evmcs->revision_id = vmcs_revision();
+	}
 
 	return true;
 }

tools/testing/selftests/kvm/x86_64/evmcs_test.c

+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2018, Red Hat, Inc.
+ *
+ * Tests for Enlightened VMCS, including nested guest state.
+ */
+#define _GNU_SOURCE /* for program_invocation_short_name */
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/ioctl.h>
+
+#include "test_util.h"
+
+#include "kvm_util.h"
+
+#include "vmx.h"
+
+#define VCPU_ID		5
+
+static bool have_nested_state;
+
+void l2_guest_code(void)
+{
+	GUEST_SYNC(6);
+
+	GUEST_SYNC(7);
+
+	/* Done, exit to L1 and never come back.  */
+	vmcall();
+}
+
+void l1_guest_code(struct vmx_pages *vmx_pages)
+{
+#define L2_GUEST_STACK_SIZE 64
+	unsigned long l2_guest_stack[L2_GUEST_STACK_SIZE];
+
+	enable_vp_assist(vmx_pages->vp_assist_gpa, vmx_pages->vp_assist);
+
+	GUEST_ASSERT(vmx_pages->vmcs_gpa);
+	GUEST_ASSERT(prepare_for_vmx_operation(vmx_pages));
+	GUEST_SYNC(3);
+	GUEST_ASSERT(load_vmcs(vmx_pages));
+	GUEST_ASSERT(vmptrstz() == vmx_pages->enlightened_vmcs_gpa);
+
+	GUEST_SYNC(4);
+	GUEST_ASSERT(vmptrstz() == vmx_pages->enlightened_vmcs_gpa);
+
+	prepare_vmcs(vmx_pages, l2_guest_code,
+		     &l2_guest_stack[L2_GUEST_STACK_SIZE]);
+
+	GUEST_SYNC(5);
+	GUEST_ASSERT(vmptrstz() == vmx_pages->enlightened_vmcs_gpa);
+	GUEST_ASSERT(!vmlaunch());
+	GUEST_ASSERT(vmptrstz() == vmx_pages->enlightened_vmcs_gpa);
+	GUEST_SYNC(8);
+	GUEST_ASSERT(!vmresume());
+	GUEST_ASSERT(vmreadz(VM_EXIT_REASON) == EXIT_REASON_VMCALL);
+	GUEST_SYNC(9);
+}
+
+void guest_code(struct vmx_pages *vmx_pages)
+{
+	GUEST_SYNC(1);
+	GUEST_SYNC(2);
+
+	if (vmx_pages)
+		l1_guest_code(vmx_pages);
+
+	GUEST_DONE();
+}
+
+int main(int argc, char *argv[])
+{
+	struct vmx_pages *vmx_pages = NULL;
+	vm_vaddr_t vmx_pages_gva = 0;
+
+	struct kvm_regs regs1, regs2;
+	struct kvm_vm *vm;
+	struct kvm_run *run;
+	struct kvm_x86_state *state;
+	struct ucall uc;
+	int stage;
+	uint16_t evmcs_ver;
+	struct kvm_enable_cap enable_evmcs_cap = {
+		.cap = KVM_CAP_HYPERV_ENLIGHTENED_VMCS,
+		 .args[0] = (unsigned long)&evmcs_ver
+	};
+
+	struct kvm_cpuid_entry2 *entry = kvm_get_supported_cpuid_entry(1);
+
+	/* Create VM */
+	vm = vm_create_default(VCPU_ID, 0, guest_code);
+
+	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
+
+	if (!kvm_check_cap(KVM_CAP_NESTED_STATE) ||
+	    !kvm_check_cap(KVM_CAP_HYPERV_ENLIGHTENED_VMCS)) {
+		printf("capabilities not available, skipping test\n");
+		exit(KSFT_SKIP);
+	}
+
+	vcpu_ioctl(vm, VCPU_ID, KVM_ENABLE_CAP, &enable_evmcs_cap);
+
+	run = vcpu_state(vm, VCPU_ID);
+
+	vcpu_regs_get(vm, VCPU_ID, &regs1);
+
+	vmx_pages = vcpu_alloc_vmx(vm, &vmx_pages_gva);
+	vcpu_args_set(vm, VCPU_ID, 1, vmx_pages_gva);
+
+	for (stage = 1;; stage++) {
+		_vcpu_run(vm, VCPU_ID);
+		TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
+			    "Unexpected exit reason: %u (%s),\n",
+			    run->exit_reason,
+			    exit_reason_str(run->exit_reason));
+
+		memset(&regs1, 0, sizeof(regs1));
+		vcpu_regs_get(vm, VCPU_ID, &regs1);
+		switch (get_ucall(vm, VCPU_ID, &uc)) {
+		case UCALL_ABORT:
+			TEST_ASSERT(false, "%s at %s:%d", (const char *)uc.args[0],
+				    __FILE__, uc.args[1]);
+			/* NOT REACHED */
+		case UCALL_SYNC:
+			break;
+		case UCALL_DONE:
+			goto done;
+		default:
+			TEST_ASSERT(false, "Unknown ucall 0x%x.", uc.cmd);
+		}
+
+		/* UCALL_SYNC is handled here.  */
+		TEST_ASSERT(!strcmp((const char *)uc.args[0], "hello") &&
+			    uc.args[1] == stage, "Unexpected register values vmexit #%lx, got %lx",
+			    stage, (ulong)uc.args[1]);
+
+		state = vcpu_save_state(vm, VCPU_ID);
+		kvm_vm_release(vm);
+
+		/* Restore state in a new VM.  */
+		kvm_vm_restart(vm, O_RDWR);
+		vm_vcpu_add(vm, VCPU_ID, 0, 0);
+		vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
+		vcpu_load_state(vm, VCPU_ID, state);
+		run = vcpu_state(vm, VCPU_ID);
+		free(state);
+
+		memset(&regs2, 0, sizeof(regs2));
+		vcpu_regs_get(vm, VCPU_ID, &regs2);
+		TEST_ASSERT(!memcmp(&regs1, &regs2, sizeof(regs2)),
+			    "Unexpected register values after vcpu_load_state; rdi: %lx rsi: %lx",
+			    (ulong) regs2.rdi, (ulong) regs2.rsi);
+	}
+
+done:
+	kvm_vm_free(vm);
+}