Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull kvm fixes from Paolo Bonzini:
 "Bug fixes, plus a new test case and the associated infrastructure for
  writing nested virtualization tests"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  kvm: selftests: add vmx_tsc_adjust_test
  kvm: x86: move MSR_IA32_TSC handling to x86.c
  X86/KVM: Properly update 'tsc_offset' to represent the running guest
  kvm: selftests: add -std=gnu99 cflags
  x86: Add check for APIC access address for vmentry of L2 guests
  KVM: X86: fix incorrect reference of trace_kvm_pi_irte_update
  X86/KVM: Do not allow DISABLE_EXITS_MWAIT when LAPIC ARAT is not available
  kvm: selftests: fix spelling mistake: "divisable" and "divisible"
  X86/VMX: Disable VMX preemption timer if MWAIT is not intercepted

arch/x86/include/asm/kvm_host.h

@@ -1013,6 +1013,7 @@ struct kvm_x86_ops {
 
 	bool (*has_wbinvd_exit)(void);
 
+	u64 (*read_l1_tsc_offset)(struct kvm_vcpu *vcpu);
 	void (*write_tsc_offset)(struct kvm_vcpu *vcpu, u64 offset);
 
 	void (*get_exit_info)(struct kvm_vcpu *vcpu, u64 *info1, u64 *info2);

arch/x86/kvm/svm.c

@@ -1423,12 +1423,23 @@ static void init_sys_seg(struct vmcb_seg *seg, uint32_t type)
 	seg->base = 0;
 }
 
+static u64 svm_read_l1_tsc_offset(struct kvm_vcpu *vcpu)
+{
+	struct vcpu_svm *svm = to_svm(vcpu);
+
+	if (is_guest_mode(vcpu))
+		return svm->nested.hsave->control.tsc_offset;
+
+	return vcpu->arch.tsc_offset;
+}
+
 static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 	u64 g_tsc_offset = 0;
 
 	if (is_guest_mode(vcpu)) {
+		/* Write L1's TSC offset.  */
 		g_tsc_offset = svm->vmcb->control.tsc_offset -
 			       svm->nested.hsave->control.tsc_offset;
 		svm->nested.hsave->control.tsc_offset = offset;
@@ -3322,6 +3333,7 @@ static int nested_svm_vmexit(struct vcpu_svm *svm)
 	/* Restore the original control entries */
 	copy_vmcb_control_area(vmcb, hsave);
 
+	svm->vcpu.arch.tsc_offset = svm->vmcb->control.tsc_offset;
 	kvm_clear_exception_queue(&svm->vcpu);
 	kvm_clear_interrupt_queue(&svm->vcpu);
 
@@ -3482,10 +3494,12 @@ static void enter_svm_guest_mode(struct vcpu_svm *svm, u64 vmcb_gpa,
 	/* We don't want to see VMMCALLs from a nested guest */
 	clr_intercept(svm, INTERCEPT_VMMCALL);
 
+	svm->vcpu.arch.tsc_offset += nested_vmcb->control.tsc_offset;
+	svm->vmcb->control.tsc_offset = svm->vcpu.arch.tsc_offset;
+
 	svm->vmcb->control.virt_ext = nested_vmcb->control.virt_ext;
 	svm->vmcb->control.int_vector = nested_vmcb->control.int_vector;
 	svm->vmcb->control.int_state = nested_vmcb->control.int_state;
-	svm->vmcb->control.tsc_offset += nested_vmcb->control.tsc_offset;
 	svm->vmcb->control.event_inj = nested_vmcb->control.event_inj;
 	svm->vmcb->control.event_inj_err = nested_vmcb->control.event_inj_err;
 
@@ -4035,12 +4049,6 @@ static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 	struct vcpu_svm *svm = to_svm(vcpu);
 
 	switch (msr_info->index) {
-	case MSR_IA32_TSC: {
-		msr_info->data = svm->vmcb->control.tsc_offset +
-			kvm_scale_tsc(vcpu, rdtsc());
-
-		break;
-	}
 	case MSR_STAR:
 		msr_info->data = svm->vmcb->save.star;
 		break;
@@ -4193,9 +4201,6 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
 		svm->vmcb->save.g_pat = data;
 		mark_dirty(svm->vmcb, VMCB_NPT);
 		break;
-	case MSR_IA32_TSC:
-		kvm_write_tsc(vcpu, msr);
-		break;
 	case MSR_IA32_SPEC_CTRL:
 		if (!msr->host_initiated &&
 		    !guest_cpuid_has(vcpu, X86_FEATURE_IBRS))
@@ -5265,9 +5270,8 @@ static int svm_update_pi_irte(struct kvm *kvm, unsigned int host_irq,
 		}
 
 		if (!ret && svm) {
-			trace_kvm_pi_irte_update(svm->vcpu.vcpu_id,
-						 host_irq, e->gsi,
-						 vcpu_info.vector,
+			trace_kvm_pi_irte_update(host_irq, svm->vcpu.vcpu_id,
+						 e->gsi, vcpu_info.vector,
 						 vcpu_info.pi_desc_addr, set);
 		}
 
@@ -7102,6 +7106,7 @@ static struct kvm_x86_ops svm_x86_ops __ro_after_init = {
 
 	.has_wbinvd_exit = svm_has_wbinvd_exit,
 
+	.read_l1_tsc_offset = svm_read_l1_tsc_offset,
 	.write_tsc_offset = svm_write_tsc_offset,
 
 	.set_tdp_cr3 = set_tdp_cr3,

arch/x86/kvm/vmx.c

@@ -2880,18 +2880,15 @@ static void setup_msrs(struct vcpu_vmx *vmx)
 		vmx_update_msr_bitmap(&vmx->vcpu);
 }
 
-/*
- * reads and returns guest's timestamp counter "register"
- * guest_tsc = (host_tsc * tsc multiplier) >> 48 + tsc_offset
- * -- Intel TSC Scaling for Virtualization White Paper, sec 1.3
- */
-static u64 guest_read_tsc(struct kvm_vcpu *vcpu)
+static u64 vmx_read_l1_tsc_offset(struct kvm_vcpu *vcpu)
 {
-	u64 host_tsc, tsc_offset;
+	struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
 
-	host_tsc = rdtsc();
-	tsc_offset = vmcs_read64(TSC_OFFSET);
-	return kvm_scale_tsc(vcpu, host_tsc) + tsc_offset;
+	if (is_guest_mode(vcpu) &&
+	    (vmcs12->cpu_based_vm_exec_control & CPU_BASED_USE_TSC_OFFSETING))
+		return vcpu->arch.tsc_offset - vmcs12->tsc_offset;
+
+	return vcpu->arch.tsc_offset;
 }
 
 /*
@@ -3524,9 +3521,6 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 #endif
 	case MSR_EFER:
 		return kvm_get_msr_common(vcpu, msr_info);
-	case MSR_IA32_TSC:
-		msr_info->data = guest_read_tsc(vcpu);
-		break;
 	case MSR_IA32_SPEC_CTRL:
 		if (!msr_info->host_initiated &&
 		    !guest_cpuid_has(vcpu, X86_FEATURE_IBRS) &&
@@ -3646,9 +3640,6 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 			return 1;
 		vmcs_write64(GUEST_BNDCFGS, data);
 		break;
-	case MSR_IA32_TSC:
-		kvm_write_tsc(vcpu, msr_info);
-		break;
 	case MSR_IA32_SPEC_CTRL:
 		if (!msr_info->host_initiated &&
 		    !guest_cpuid_has(vcpu, X86_FEATURE_IBRS) &&
@@ -10608,6 +10599,16 @@ static inline bool nested_vmx_prepare_msr_bitmap(struct kvm_vcpu *vcpu,
 	return true;
 }
 
+static int nested_vmx_check_apic_access_controls(struct kvm_vcpu *vcpu,
+					  struct vmcs12 *vmcs12)
+{
+	if (nested_cpu_has2(vmcs12, SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES) &&
+	    !page_address_valid(vcpu, vmcs12->apic_access_addr))
+		return -EINVAL;
+	else
+		return 0;
+}
+
 static int nested_vmx_check_apicv_controls(struct kvm_vcpu *vcpu,
 					   struct vmcs12 *vmcs12)
 {
@@ -11176,11 +11177,8 @@ static int prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12,
 		vmcs_write64(GUEST_IA32_PAT, vmx->vcpu.arch.pat);
 	}
 
-	if (vmcs12->cpu_based_vm_exec_control & CPU_BASED_USE_TSC_OFFSETING)
-		vmcs_write64(TSC_OFFSET,
-			vcpu->arch.tsc_offset + vmcs12->tsc_offset);
-	else
 	vmcs_write64(TSC_OFFSET, vcpu->arch.tsc_offset);
+
 	if (kvm_has_tsc_control)
 		decache_tsc_multiplier(vmx);
 
@@ -11299,6 +11297,9 @@ static int check_vmentry_prereqs(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
 	if (nested_vmx_check_msr_bitmap_controls(vcpu, vmcs12))
 		return VMXERR_ENTRY_INVALID_CONTROL_FIELD;
 
+	if (nested_vmx_check_apic_access_controls(vcpu, vmcs12))
+		return VMXERR_ENTRY_INVALID_CONTROL_FIELD;
+
 	if (nested_vmx_check_tpr_shadow_controls(vcpu, vmcs12))
 		return VMXERR_ENTRY_INVALID_CONTROL_FIELD;
 
@@ -11420,6 +11421,7 @@ static int enter_vmx_non_root_mode(struct kvm_vcpu *vcpu, bool from_vmentry)
 	struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
 	u32 msr_entry_idx;
 	u32 exit_qual;
+	int r;
 
 	enter_guest_mode(vcpu);
 
@@ -11429,26 +11431,21 @@ static int enter_vmx_non_root_mode(struct kvm_vcpu *vcpu, bool from_vmentry)
 	vmx_switch_vmcs(vcpu, &vmx->nested.vmcs02);
 	vmx_segment_cache_clear(vmx);
 
-	if (prepare_vmcs02(vcpu, vmcs12, from_vmentry, &exit_qual)) {
-		leave_guest_mode(vcpu);
-		vmx_switch_vmcs(vcpu, &vmx->vmcs01);
-		nested_vmx_entry_failure(vcpu, vmcs12,
-					 EXIT_REASON_INVALID_STATE, exit_qual);
-		return 1;
-	}
+	if (vmcs12->cpu_based_vm_exec_control & CPU_BASED_USE_TSC_OFFSETING)
+		vcpu->arch.tsc_offset += vmcs12->tsc_offset;
+
+	r = EXIT_REASON_INVALID_STATE;
+	if (prepare_vmcs02(vcpu, vmcs12, from_vmentry, &exit_qual))
+		goto fail;
 
 	nested_get_vmcs12_pages(vcpu, vmcs12);
 
+	r = EXIT_REASON_MSR_LOAD_FAIL;
 	msr_entry_idx = nested_vmx_load_msr(vcpu,
 					    vmcs12->vm_entry_msr_load_addr,
 					    vmcs12->vm_entry_msr_load_count);
-	if (msr_entry_idx) {
-		leave_guest_mode(vcpu);
-		vmx_switch_vmcs(vcpu, &vmx->vmcs01);
-		nested_vmx_entry_failure(vcpu, vmcs12,
-				EXIT_REASON_MSR_LOAD_FAIL, msr_entry_idx);
-		return 1;
-	}
+	if (msr_entry_idx)
+		goto fail;
 
 	/*
 	 * Note no nested_vmx_succeed or nested_vmx_fail here. At this point
@@ -11457,6 +11454,14 @@ static int enter_vmx_non_root_mode(struct kvm_vcpu *vcpu, bool from_vmentry)
 	 * the success flag) when L2 exits (see nested_vmx_vmexit()).
 	 */
 	return 0;
+
+fail:
+	if (vmcs12->cpu_based_vm_exec_control & CPU_BASED_USE_TSC_OFFSETING)
+		vcpu->arch.tsc_offset -= vmcs12->tsc_offset;
+	leave_guest_mode(vcpu);
+	vmx_switch_vmcs(vcpu, &vmx->vmcs01);
+	nested_vmx_entry_failure(vcpu, vmcs12, r, exit_qual);
+	return 1;
 }
 
 /*
@@ -12028,6 +12033,9 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason,
 
 	leave_guest_mode(vcpu);
 
+	if (vmcs12->cpu_based_vm_exec_control & CPU_BASED_USE_TSC_OFFSETING)
+		vcpu->arch.tsc_offset -= vmcs12->tsc_offset;
+
 	if (likely(!vmx->fail)) {
 		if (exit_reason == -1)
 			sync_vmcs12(vcpu, vmcs12);
@@ -12224,10 +12232,16 @@ static inline int u64_shl_div_u64(u64 a, unsigned int shift,
 
 static int vmx_set_hv_timer(struct kvm_vcpu *vcpu, u64 guest_deadline_tsc)
 {
-	struct vcpu_vmx *vmx = to_vmx(vcpu);
-	u64 tscl = rdtsc();
-	u64 guest_tscl = kvm_read_l1_tsc(vcpu, tscl);
-	u64 delta_tsc = max(guest_deadline_tsc, guest_tscl) - guest_tscl;
+	struct vcpu_vmx *vmx;
+	u64 tscl, guest_tscl, delta_tsc;
+
+	if (kvm_mwait_in_guest(vcpu->kvm))
+		return -EOPNOTSUPP;
+
+	vmx = to_vmx(vcpu);
+	tscl = rdtsc();
+	guest_tscl = kvm_read_l1_tsc(vcpu, tscl);
+	delta_tsc = max(guest_deadline_tsc, guest_tscl) - guest_tscl;
 
 	/* Convert to host delta tsc if tsc scaling is enabled */
 	if (vcpu->arch.tsc_scaling_ratio != kvm_default_tsc_scaling_ratio &&
@@ -12533,7 +12547,7 @@ static int vmx_update_pi_irte(struct kvm *kvm, unsigned int host_irq,
 		vcpu_info.pi_desc_addr = __pa(vcpu_to_pi_desc(vcpu));
 		vcpu_info.vector = irq.vector;
 
-		trace_kvm_pi_irte_update(vcpu->vcpu_id, host_irq, e->gsi,
+		trace_kvm_pi_irte_update(host_irq, vcpu->vcpu_id, e->gsi,
 				vcpu_info.vector, vcpu_info.pi_desc_addr, set);
 
 		if (set)
@@ -12712,6 +12726,7 @@ static struct kvm_x86_ops vmx_x86_ops __ro_after_init = {
 
 	.has_wbinvd_exit = cpu_has_vmx_wbinvd_exit,
 
+	.read_l1_tsc_offset = vmx_read_l1_tsc_offset,
 	.write_tsc_offset = vmx_write_tsc_offset,
 
 	.set_tdp_cr3 = vmx_set_cr3,

arch/x86/kvm/x86.c

@@ -1490,7 +1490,7 @@ static void kvm_track_tsc_matching(struct kvm_vcpu *vcpu)
 
 static void update_ia32_tsc_adjust_msr(struct kvm_vcpu *vcpu, s64 offset)
 {
-	u64 curr_offset = vcpu->arch.tsc_offset;
+	u64 curr_offset = kvm_x86_ops->read_l1_tsc_offset(vcpu);
 	vcpu->arch.ia32_tsc_adjust_msr += offset - curr_offset;
 }
 
@@ -1532,7 +1532,9 @@ static u64 kvm_compute_tsc_offset(struct kvm_vcpu *vcpu, u64 target_tsc)
 
 u64 kvm_read_l1_tsc(struct kvm_vcpu *vcpu, u64 host_tsc)
 {
-	return vcpu->arch.tsc_offset + kvm_scale_tsc(vcpu, host_tsc);
+	u64 tsc_offset = kvm_x86_ops->read_l1_tsc_offset(vcpu);
+
+	return tsc_offset + kvm_scale_tsc(vcpu, host_tsc);
 }
 EXPORT_SYMBOL_GPL(kvm_read_l1_tsc);
 
@@ -2362,6 +2364,9 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 			return 1;
 		vcpu->arch.smbase = data;
 		break;
+	case MSR_IA32_TSC:
+		kvm_write_tsc(vcpu, msr_info);
+		break;
 	case MSR_SMI_COUNT:
 		if (!msr_info->host_initiated)
 			return 1;
@@ -2605,6 +2610,9 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 	case MSR_IA32_UCODE_REV:
 		msr_info->data = vcpu->arch.microcode_version;
 		break;
+	case MSR_IA32_TSC:
+		msr_info->data = kvm_scale_tsc(vcpu, rdtsc()) + vcpu->arch.tsc_offset;
+		break;
 	case MSR_MTRRcap:
 	case 0x200 ... 0x2ff:
 		return kvm_mtrr_get_msr(vcpu, msr_info->index, &msr_info->data);
@@ -2819,7 +2827,8 @@ static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs,
 static inline bool kvm_can_mwait_in_guest(void)
 {
 	return boot_cpu_has(X86_FEATURE_MWAIT) &&
-		!boot_cpu_has_bug(X86_BUG_MONITOR);
+		!boot_cpu_has_bug(X86_BUG_MONITOR) &&
+		boot_cpu_has(X86_FEATURE_ARAT);
 }
 
 int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)

tools/testing/selftests/kvm/Makefile

@@ -4,17 +4,18 @@ top_srcdir = ../../../../
 UNAME_M := $(shell uname -m)
 
 LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c
-LIBKVM_x86_64 = lib/x86.c
+LIBKVM_x86_64 = lib/x86.c lib/vmx.c
 
 TEST_GEN_PROGS_x86_64 = set_sregs_test
 TEST_GEN_PROGS_x86_64 += sync_regs_test
+TEST_GEN_PROGS_x86_64 += vmx_tsc_adjust_test
 
 TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(UNAME_M))
 LIBKVM += $(LIBKVM_$(UNAME_M))
 
 INSTALL_HDR_PATH = $(top_srcdir)/usr
 LINUX_HDR_PATH = $(INSTALL_HDR_PATH)/include/
-CFLAGS += -O2 -g -I$(LINUX_HDR_PATH) -Iinclude -I$(<D)
+CFLAGS += -O2 -g -std=gnu99 -I$(LINUX_HDR_PATH) -Iinclude -I$(<D)
 
 # After inclusion, $(OUTPUT) is defined and
 # $(TEST_GEN_PROGS) starts with $(OUTPUT)/

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

@@ -112,24 +112,27 @@ void virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
 vm_paddr_t vm_phy_page_alloc(struct kvm_vm *vm,
 	vm_paddr_t paddr_min, uint32_t memslot);
 
-void kvm_get_supported_cpuid(struct kvm_cpuid2 *cpuid);
+struct kvm_cpuid2 *kvm_get_supported_cpuid(void);
 void vcpu_set_cpuid(
 	struct kvm_vm *vm, uint32_t vcpuid, struct kvm_cpuid2 *cpuid);
 
-struct kvm_cpuid2 *allocate_kvm_cpuid2(void);
 struct kvm_cpuid_entry2 *
-find_cpuid_index_entry(struct kvm_cpuid2 *cpuid, uint32_t function,
-		       uint32_t index);
+kvm_get_supported_cpuid_index(uint32_t function, uint32_t index);
 
 static inline struct kvm_cpuid_entry2 *
-find_cpuid_entry(struct kvm_cpuid2 *cpuid, uint32_t function)
+kvm_get_supported_cpuid_entry(uint32_t function)
 {
-	return find_cpuid_index_entry(cpuid, function, 0);
+	return kvm_get_supported_cpuid_index(function, 0);
 }
 
 struct kvm_vm *vm_create_default(uint32_t vcpuid, void *guest_code);
 void vm_vcpu_add_default(struct kvm_vm *vm, uint32_t vcpuid, void *guest_code);
 
+typedef void (*vmx_guest_code_t)(vm_vaddr_t vmxon_vaddr,
+				 vm_paddr_t vmxon_paddr,
+				 vm_vaddr_t vmcs_vaddr,
+				 vm_paddr_t vmcs_paddr);
+
 struct kvm_userspace_memory_region *
 kvm_userspace_memory_region_find(struct kvm_vm *vm, uint64_t start,
 				 uint64_t end);

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

@@ -378,7 +378,7 @@ int kvm_memcmp_hva_gva(void *hva,
  * complicated. This function uses a reasonable default length for
  * the array and performs the appropriate allocation.
  */
-struct kvm_cpuid2 *allocate_kvm_cpuid2(void)
+static struct kvm_cpuid2 *allocate_kvm_cpuid2(void)
 {
 	struct kvm_cpuid2 *cpuid;
 	int nent = 100;
@@ -402,17 +402,21 @@ struct kvm_cpuid2 *allocate_kvm_cpuid2(void)
  * Input Args: None
  *
  * Output Args:
- *   cpuid - The supported KVM CPUID
  *
- * Return: void
+ * Return: The supported KVM CPUID
  *
  * Get the guest CPUID supported by KVM.
  */
-void kvm_get_supported_cpuid(struct kvm_cpuid2 *cpuid)
+struct kvm_cpuid2 *kvm_get_supported_cpuid(void)
 {
+	static struct kvm_cpuid2 *cpuid;
 	int ret;
 	int kvm_fd;
 
+	if (cpuid)
+		return cpuid;
+
+	cpuid = allocate_kvm_cpuid2();
 	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
 	TEST_ASSERT(kvm_fd >= 0, "open %s failed, rc: %i errno: %i",
 		KVM_DEV_PATH, kvm_fd, errno);
@@ -422,6 +426,7 @@ void kvm_get_supported_cpuid(struct kvm_cpuid2 *cpuid)
 		    ret, errno);
 
 	close(kvm_fd);
+	return cpuid;
 }
 
 /* Locate a cpuid entry.
@@ -435,12 +440,13 @@ void kvm_get_supported_cpuid(struct kvm_cpuid2 *cpuid)
  * Return: A pointer to the cpuid entry. Never returns NULL.
  */
 struct kvm_cpuid_entry2 *
-find_cpuid_index_entry(struct kvm_cpuid2 *cpuid, uint32_t function,
-		       uint32_t index)
+kvm_get_supported_cpuid_index(uint32_t function, uint32_t index)
 {
+	struct kvm_cpuid2 *cpuid;
 	struct kvm_cpuid_entry2 *entry = NULL;
 	int i;
 
+	cpuid = kvm_get_supported_cpuid();
 	for (i = 0; i < cpuid->nent; i++) {
 		if (cpuid->entries[i].function == function &&
 		    cpuid->entries[i].index == index) {
@@ -1435,7 +1441,7 @@ vm_paddr_t vm_phy_page_alloc(struct kvm_vm *vm,
 	sparsebit_idx_t pg;
 
 	TEST_ASSERT((paddr_min % vm->page_size) == 0, "Min physical address "
-		"not divisable by page size.\n"
+		"not divisible by page size.\n"
 		"  paddr_min: 0x%lx page_size: 0x%x",
 		paddr_min, vm->page_size);
 

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

@@ -121,7 +121,7 @@
  *     avoided by moving the setting of the nodes mask bits into
  *     the previous nodes num_after setting.
  *
- *   + Node starting index is evenly divisable by the number of bits
+ *   + Node starting index is evenly divisible by the number of bits
  *     within a nodes mask member.
  *
  *   + Nodes never represent a range of bits that wrap around the
@@ -1741,7 +1741,7 @@ void sparsebit_validate_internal(struct sparsebit *s)
 
 		/* Validate node index is divisible by the mask size */
 		if (nodep->idx % MASK_BITS) {
-			fprintf(stderr, "Node index not divisable by "
+			fprintf(stderr, "Node index not divisible by "
 				"mask size,\n"
 				"  nodep: %p nodep->idx: 0x%lx "
 				"MASK_BITS: %lu\n",

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

+/*
+ * tools/testing/selftests/kvm/lib/x86.c
+ *
+ * Copyright (C) 2018, Google LLC.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ */
+
+#define _GNU_SOURCE /* for program_invocation_name */
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "x86.h"
+#include "vmx.h"
+
+/* Create a default VM for VMX tests.
+ *
+ * Input Args:
+ *   vcpuid - The id of the single VCPU to add to the VM.
+ *   guest_code - The vCPU's entry point
+ *
+ * Output Args: None
+ *
+ * Return:
+ *   Pointer to opaque structure that describes the created VM.
+ */
+struct kvm_vm *
+vm_create_default_vmx(uint32_t vcpuid, vmx_guest_code_t guest_code)
+{
+	struct kvm_cpuid2 *cpuid;
+	struct kvm_vm *vm;
+	vm_vaddr_t vmxon_vaddr;
+	vm_paddr_t vmxon_paddr;
+	vm_vaddr_t vmcs_vaddr;
+	vm_paddr_t vmcs_paddr;
+
+	vm = vm_create_default(vcpuid, (void *) guest_code);
+
+	/* Enable nesting in CPUID */
+	vcpu_set_cpuid(vm, vcpuid, kvm_get_supported_cpuid());
+
+	/* Setup of a region of guest memory for the vmxon region. */
+	vmxon_vaddr = vm_vaddr_alloc(vm, getpagesize(), 0, 0, 0);
+	vmxon_paddr = addr_gva2gpa(vm, vmxon_vaddr);
+
+	/* Setup of a region of guest memory for a vmcs. */
+	vmcs_vaddr = vm_vaddr_alloc(vm, getpagesize(), 0, 0, 0);
+	vmcs_paddr = addr_gva2gpa(vm, vmcs_vaddr);
+
+	vcpu_args_set(vm, vcpuid, 4, vmxon_vaddr, vmxon_paddr, vmcs_vaddr,
+		      vmcs_paddr);
+
+	return vm;
+}
+
+void prepare_for_vmx_operation(void)
+{
+	uint64_t feature_control;
+	uint64_t required;
+	unsigned long cr0;
+	unsigned long cr4;
+
+	/*
+	 * Ensure bits in CR0 and CR4 are valid in VMX operation:
+	 * - Bit X is 1 in _FIXED0: bit X is fixed to 1 in CRx.
+	 * - Bit X is 0 in _FIXED1: bit X is fixed to 0 in CRx.
+	 */
+	__asm__ __volatile__("mov %%cr0, %0" : "=r"(cr0) : : "memory");
+	cr0 &= rdmsr(MSR_IA32_VMX_CR0_FIXED1);
+	cr0 |= rdmsr(MSR_IA32_VMX_CR0_FIXED0);
+	__asm__ __volatile__("mov %0, %%cr0" : : "r"(cr0) : "memory");
+
+	__asm__ __volatile__("mov %%cr4, %0" : "=r"(cr4) : : "memory");
+	cr4 &= rdmsr(MSR_IA32_VMX_CR4_FIXED1);
+	cr4 |= rdmsr(MSR_IA32_VMX_CR4_FIXED0);
+	/* Enable VMX operation */
+	cr4 |= X86_CR4_VMXE;
+	__asm__ __volatile__("mov %0, %%cr4" : : "r"(cr4) : "memory");
+
+	/*
+	 * Configure IA32_FEATURE_CONTROL MSR to allow VMXON:
+	 *  Bit 0: Lock bit. If clear, VMXON causes a #GP.
+	 *  Bit 2: Enables VMXON outside of SMX operation. If clear, VMXON
+	 *    outside of SMX causes a #GP.
+	 */
+	required = FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX;
+	required |= FEATURE_CONTROL_LOCKED;
+	feature_control = rdmsr(MSR_IA32_FEATURE_CONTROL);
+	if ((feature_control & required) != required)
+		wrmsr(MSR_IA32_FEATURE_CONTROL, feature_control | required);
+}
+
+/*
+ * Initialize the control fields to the most basic settings possible.
+ */
+static inline void init_vmcs_control_fields(void)
+{
+	vmwrite(VIRTUAL_PROCESSOR_ID, 0);
+	vmwrite(POSTED_INTR_NV, 0);
+
+	vmwrite(PIN_BASED_VM_EXEC_CONTROL, rdmsr(MSR_IA32_VMX_PINBASED_CTLS));
+	vmwrite(CPU_BASED_VM_EXEC_CONTROL, rdmsr(MSR_IA32_VMX_PROCBASED_CTLS));
+	vmwrite(EXCEPTION_BITMAP, 0);
+	vmwrite(PAGE_FAULT_ERROR_CODE_MASK, 0);
+	vmwrite(PAGE_FAULT_ERROR_CODE_MATCH, -1); /* Never match */
+	vmwrite(CR3_TARGET_COUNT, 0);
+	vmwrite(VM_EXIT_CONTROLS, rdmsr(MSR_IA32_VMX_EXIT_CTLS) |
+		VM_EXIT_HOST_ADDR_SPACE_SIZE);	  /* 64-bit host */
+	vmwrite(VM_EXIT_MSR_STORE_COUNT, 0);
+	vmwrite(VM_EXIT_MSR_LOAD_COUNT, 0);
+	vmwrite(VM_ENTRY_CONTROLS, rdmsr(MSR_IA32_VMX_ENTRY_CTLS) |
+		VM_ENTRY_IA32E_MODE);		  /* 64-bit guest */
+	vmwrite(VM_ENTRY_MSR_LOAD_COUNT, 0);
+	vmwrite(VM_ENTRY_INTR_INFO_FIELD, 0);
+	vmwrite(TPR_THRESHOLD, 0);
+	vmwrite(SECONDARY_VM_EXEC_CONTROL, 0);
+
+	vmwrite(CR0_GUEST_HOST_MASK, 0);
+	vmwrite(CR4_GUEST_HOST_MASK, 0);
+	vmwrite(CR0_READ_SHADOW, get_cr0());
+	vmwrite(CR4_READ_SHADOW, get_cr4());
+}
+
+/*
+ * Initialize the host state fields based on the current host state, with
+ * the exception of HOST_RSP and HOST_RIP, which should be set by vmlaunch
+ * or vmresume.
+ */
+static inline void init_vmcs_host_state(void)
+{
+	uint32_t exit_controls = vmreadz(VM_EXIT_CONTROLS);
+
+	vmwrite(HOST_ES_SELECTOR, get_es());
+	vmwrite(HOST_CS_SELECTOR, get_cs());
+	vmwrite(HOST_SS_SELECTOR, get_ss());
+	vmwrite(HOST_DS_SELECTOR, get_ds());
+	vmwrite(HOST_FS_SELECTOR, get_fs());
+	vmwrite(HOST_GS_SELECTOR, get_gs());
+	vmwrite(HOST_TR_SELECTOR, get_tr());
+
+	if (exit_controls & VM_EXIT_LOAD_IA32_PAT)
+		vmwrite(HOST_IA32_PAT, rdmsr(MSR_IA32_CR_PAT));
+	if (exit_controls & VM_EXIT_LOAD_IA32_EFER)
+		vmwrite(HOST_IA32_EFER, rdmsr(MSR_EFER));
+	if (exit_controls & VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL)
+		vmwrite(HOST_IA32_PERF_GLOBAL_CTRL,
+			rdmsr(MSR_CORE_PERF_GLOBAL_CTRL));
+
+	vmwrite(HOST_IA32_SYSENTER_CS, rdmsr(MSR_IA32_SYSENTER_CS));
+
+	vmwrite(HOST_CR0, get_cr0());
+	vmwrite(HOST_CR3, get_cr3());
+	vmwrite(HOST_CR4, get_cr4());
+	vmwrite(HOST_FS_BASE, rdmsr(MSR_FS_BASE));
+	vmwrite(HOST_GS_BASE, rdmsr(MSR_GS_BASE));
+	vmwrite(HOST_TR_BASE,
+		get_desc64_base((struct desc64 *)(get_gdt_base() + get_tr())));
+	vmwrite(HOST_GDTR_BASE, get_gdt_base());
+	vmwrite(HOST_IDTR_BASE, get_idt_base());
+	vmwrite(HOST_IA32_SYSENTER_ESP, rdmsr(MSR_IA32_SYSENTER_ESP));
+	vmwrite(HOST_IA32_SYSENTER_EIP, rdmsr(MSR_IA32_SYSENTER_EIP));
+}
+
+/*
+ * Initialize the guest state fields essentially as a clone of
+ * the host state fields. Some host state fields have fixed
+ * values, and we set the corresponding guest state fields accordingly.
+ */
+static inline void init_vmcs_guest_state(void *rip, void *rsp)
+{
+	vmwrite(GUEST_ES_SELECTOR, vmreadz(HOST_ES_SELECTOR));
+	vmwrite(GUEST_CS_SELECTOR, vmreadz(HOST_CS_SELECTOR));
+	vmwrite(GUEST_SS_SELECTOR, vmreadz(HOST_SS_SELECTOR));
+	vmwrite(GUEST_DS_SELECTOR, vmreadz(HOST_DS_SELECTOR));
+	vmwrite(GUEST_FS_SELECTOR, vmreadz(HOST_FS_SELECTOR));
+	vmwrite(GUEST_GS_SELECTOR, vmreadz(HOST_GS_SELECTOR));
+	vmwrite(GUEST_LDTR_SELECTOR, 0);
+	vmwrite(GUEST_TR_SELECTOR, vmreadz(HOST_TR_SELECTOR));
+	vmwrite(GUEST_INTR_STATUS, 0);
+	vmwrite(GUEST_PML_INDEX, 0);
+
+	vmwrite(VMCS_LINK_POINTER, -1ll);
+	vmwrite(GUEST_IA32_DEBUGCTL, 0);
+	vmwrite(GUEST_IA32_PAT, vmreadz(HOST_IA32_PAT));
+	vmwrite(GUEST_IA32_EFER, vmreadz(HOST_IA32_EFER));
+	vmwrite(GUEST_IA32_PERF_GLOBAL_CTRL,
+		vmreadz(HOST_IA32_PERF_GLOBAL_CTRL));
+
+	vmwrite(GUEST_ES_LIMIT, -1);
+	vmwrite(GUEST_CS_LIMIT, -1);
+	vmwrite(GUEST_SS_LIMIT, -1);
+	vmwrite(GUEST_DS_LIMIT, -1);
+	vmwrite(GUEST_FS_LIMIT, -1);
+	vmwrite(GUEST_GS_LIMIT, -1);
+	vmwrite(GUEST_LDTR_LIMIT, -1);
+	vmwrite(GUEST_TR_LIMIT, 0x67);
+	vmwrite(GUEST_GDTR_LIMIT, 0xffff);
+	vmwrite(GUEST_IDTR_LIMIT, 0xffff);
+	vmwrite(GUEST_ES_AR_BYTES,
+		vmreadz(GUEST_ES_SELECTOR) == 0 ? 0x10000 : 0xc093);
+	vmwrite(GUEST_CS_AR_BYTES, 0xa09b);
+	vmwrite(GUEST_SS_AR_BYTES, 0xc093);
+	vmwrite(GUEST_DS_AR_BYTES,
+		vmreadz(GUEST_DS_SELECTOR) == 0 ? 0x10000 : 0xc093);
+	vmwrite(GUEST_FS_AR_BYTES,
+		vmreadz(GUEST_FS_SELECTOR) == 0 ? 0x10000 : 0xc093);
+	vmwrite(GUEST_GS_AR_BYTES,
+		vmreadz(GUEST_GS_SELECTOR) == 0 ? 0x10000 : 0xc093);
+	vmwrite(GUEST_LDTR_AR_BYTES, 0x10000);
+	vmwrite(GUEST_TR_AR_BYTES, 0x8b);
+	vmwrite(GUEST_INTERRUPTIBILITY_INFO, 0);
+	vmwrite(GUEST_ACTIVITY_STATE, 0);
+	vmwrite(GUEST_SYSENTER_CS, vmreadz(HOST_IA32_SYSENTER_CS));
+	vmwrite(VMX_PREEMPTION_TIMER_VALUE, 0);
+
+	vmwrite(GUEST_CR0, vmreadz(HOST_CR0));
+	vmwrite(GUEST_CR3, vmreadz(HOST_CR3));
+	vmwrite(GUEST_CR4, vmreadz(HOST_CR4));
+	vmwrite(GUEST_ES_BASE, 0);
+	vmwrite(GUEST_CS_BASE, 0);
+	vmwrite(GUEST_SS_BASE, 0);
+	vmwrite(GUEST_DS_BASE, 0);
+	vmwrite(GUEST_FS_BASE, vmreadz(HOST_FS_BASE));
+	vmwrite(GUEST_GS_BASE, vmreadz(HOST_GS_BASE));
+	vmwrite(GUEST_LDTR_BASE, 0);
+	vmwrite(GUEST_TR_BASE, vmreadz(HOST_TR_BASE));
+	vmwrite(GUEST_GDTR_BASE, vmreadz(HOST_GDTR_BASE));
+	vmwrite(GUEST_IDTR_BASE, vmreadz(HOST_IDTR_BASE));
+	vmwrite(GUEST_DR7, 0x400);
+	vmwrite(GUEST_RSP, (uint64_t)rsp);
+	vmwrite(GUEST_RIP, (uint64_t)rip);
+	vmwrite(GUEST_RFLAGS, 2);
+	vmwrite(GUEST_PENDING_DBG_EXCEPTIONS, 0);
+	vmwrite(GUEST_SYSENTER_ESP, vmreadz(HOST_IA32_SYSENTER_ESP));
+	vmwrite(GUEST_SYSENTER_EIP, vmreadz(HOST_IA32_SYSENTER_EIP));
+}
+
+void prepare_vmcs(void *guest_rip, void *guest_rsp)
+{
+	init_vmcs_control_fields();
+	init_vmcs_host_state();
+	init_vmcs_guest_state(guest_rip, guest_rsp);
+}

tools/testing/selftests/kvm/vmx_tsc_adjust_test.c

+/*
+ * gtests/tests/vmx_tsc_adjust_test.c
+ *
+ * Copyright (C) 2018, Google LLC.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ *
+ *
+ * IA32_TSC_ADJUST test
+ *
+ * According to the SDM, "if an execution of WRMSR to the
+ * IA32_TIME_STAMP_COUNTER MSR adds (or subtracts) value X from the TSC,
+ * the logical processor also adds (or subtracts) value X from the
+ * IA32_TSC_ADJUST MSR.
+ *
+ * Note that when L1 doesn't intercept writes to IA32_TSC, a
+ * WRMSR(IA32_TSC) from L2 sets L1's TSC value, not L2's perceived TSC
+ * value.
+ *
+ * This test verifies that this unusual case is handled correctly.
+ */
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "x86.h"
+#include "vmx.h"
+
+#include <string.h>
+#include <sys/ioctl.h>
+
+#ifndef MSR_IA32_TSC_ADJUST
+#define MSR_IA32_TSC_ADJUST 0x3b
+#endif
+
+#define PAGE_SIZE	4096
+#define VCPU_ID		5
+
+#define TSC_ADJUST_VALUE (1ll << 32)
+#define TSC_OFFSET_VALUE -(1ll << 48)
+
+enum {
+	PORT_ABORT = 0x1000,
+	PORT_REPORT,
+	PORT_DONE,
+};
+
+struct vmx_page {
+	vm_vaddr_t virt;
+	vm_paddr_t phys;
+};
+
+enum {
+	VMXON_PAGE = 0,
+	VMCS_PAGE,
+	MSR_BITMAP_PAGE,
+
+	NUM_VMX_PAGES,
+};
+
+struct kvm_single_msr {
+	struct kvm_msrs header;
+	struct kvm_msr_entry entry;
+} __attribute__((packed));
+
+/* The virtual machine object. */
+static struct kvm_vm *vm;
+
+/* Array of vmx_page descriptors that is shared with the guest. */
+struct vmx_page *vmx_pages;
+
+#define exit_to_l0(_port, _arg) do_exit_to_l0(_port, (unsigned long) (_arg))
+static void do_exit_to_l0(uint16_t port, unsigned long arg)
+{
+	__asm__ __volatile__("in %[port], %%al"
+		:
+		: [port]"d"(port), "D"(arg)
+		: "rax");
+}
+
+
+#define GUEST_ASSERT(_condition) do {					     \
+	if (!(_condition))						     \
+		exit_to_l0(PORT_ABORT, "Failed guest assert: " #_condition); \
+} while (0)
+
+static void check_ia32_tsc_adjust(int64_t max)
+{
+	int64_t adjust;
+
+	adjust = rdmsr(MSR_IA32_TSC_ADJUST);
+	exit_to_l0(PORT_REPORT, adjust);
+	GUEST_ASSERT(adjust <= max);
+}
+
+static void l2_guest_code(void)
+{
+	uint64_t l1_tsc = rdtsc() - TSC_OFFSET_VALUE;
+
+	wrmsr(MSR_IA32_TSC, l1_tsc - TSC_ADJUST_VALUE);
+	check_ia32_tsc_adjust(-2 * TSC_ADJUST_VALUE);
+
+	/* Exit to L1 */
+	__asm__ __volatile__("vmcall");
+}
+
+static void l1_guest_code(struct vmx_page *vmx_pages)
+{
+#define L2_GUEST_STACK_SIZE 64
+	unsigned long l2_guest_stack[L2_GUEST_STACK_SIZE];
+	uint32_t control;
+	uintptr_t save_cr3;
+
+	GUEST_ASSERT(rdtsc() < TSC_ADJUST_VALUE);
+	wrmsr(MSR_IA32_TSC, rdtsc() - TSC_ADJUST_VALUE);
+	check_ia32_tsc_adjust(-1 * TSC_ADJUST_VALUE);
+
+	prepare_for_vmx_operation();
+
+	/* Enter VMX root operation. */
+	*(uint32_t *)vmx_pages[VMXON_PAGE].virt = vmcs_revision();
+	GUEST_ASSERT(!vmxon(vmx_pages[VMXON_PAGE].phys));
+
+	/* Load a VMCS. */
+	*(uint32_t *)vmx_pages[VMCS_PAGE].virt = vmcs_revision();
+	GUEST_ASSERT(!vmclear(vmx_pages[VMCS_PAGE].phys));
+	GUEST_ASSERT(!vmptrld(vmx_pages[VMCS_PAGE].phys));
+
+	/* Prepare the VMCS for L2 execution. */
+	prepare_vmcs(l2_guest_code, &l2_guest_stack[L2_GUEST_STACK_SIZE]);
+	control = vmreadz(CPU_BASED_VM_EXEC_CONTROL);
+	control |= CPU_BASED_USE_MSR_BITMAPS | CPU_BASED_USE_TSC_OFFSETING;
+	vmwrite(CPU_BASED_VM_EXEC_CONTROL, control);
+	vmwrite(MSR_BITMAP, vmx_pages[MSR_BITMAP_PAGE].phys);
+	vmwrite(TSC_OFFSET, TSC_OFFSET_VALUE);
+
+	/* Jump into L2.  First, test failure to load guest CR3.  */
+	save_cr3 = vmreadz(GUEST_CR3);
+	vmwrite(GUEST_CR3, -1ull);
+	GUEST_ASSERT(!vmlaunch());
+	GUEST_ASSERT(vmreadz(VM_EXIT_REASON) ==
+		     (EXIT_REASON_FAILED_VMENTRY | EXIT_REASON_INVALID_STATE));
+	check_ia32_tsc_adjust(-1 * TSC_ADJUST_VALUE);
+	vmwrite(GUEST_CR3, save_cr3);
+
+	GUEST_ASSERT(!vmlaunch());
+	GUEST_ASSERT(vmreadz(VM_EXIT_REASON) == EXIT_REASON_VMCALL);
+
+	check_ia32_tsc_adjust(-2 * TSC_ADJUST_VALUE);
+
+	exit_to_l0(PORT_DONE, 0);
+}
+
+static void allocate_vmx_page(struct vmx_page *page)
+{
+	vm_vaddr_t virt;
+
+	virt = vm_vaddr_alloc(vm, PAGE_SIZE, 0, 0, 0);
+	memset(addr_gva2hva(vm, virt), 0, PAGE_SIZE);
+
+	page->virt = virt;
+	page->phys = addr_gva2gpa(vm, virt);
+}
+
+static vm_vaddr_t allocate_vmx_pages(void)
+{
+	vm_vaddr_t vmx_pages_vaddr;
+	int i;
+
+	vmx_pages_vaddr = vm_vaddr_alloc(
+		vm, sizeof(struct vmx_page) * NUM_VMX_PAGES, 0, 0, 0);
+
+	vmx_pages = (void *) addr_gva2hva(vm, vmx_pages_vaddr);
+
+	for (i = 0; i < NUM_VMX_PAGES; i++)
+		allocate_vmx_page(&vmx_pages[i]);
+
+	return vmx_pages_vaddr;
+}
+
+void report(int64_t val)
+{
+	printf("IA32_TSC_ADJUST is %ld (%lld * TSC_ADJUST_VALUE + %lld).\n",
+	       val, val / TSC_ADJUST_VALUE, val % TSC_ADJUST_VALUE);
+}
+
+int main(int argc, char *argv[])
+{
+	vm_vaddr_t vmx_pages_vaddr;
+	struct kvm_cpuid_entry2 *entry = kvm_get_supported_cpuid_entry(1);
+
+	if (!(entry->ecx & CPUID_VMX)) {
+		printf("nested VMX not enabled, skipping test");
+		return 0;
+	}
+
+	vm = vm_create_default_vmx(VCPU_ID, (void *) l1_guest_code);
+
+	/* Allocate VMX pages and shared descriptors (vmx_pages). */
+	vmx_pages_vaddr = allocate_vmx_pages();
+	vcpu_args_set(vm, VCPU_ID, 1, vmx_pages_vaddr);
+
+	for (;;) {
+		volatile struct kvm_run *run = vcpu_state(vm, VCPU_ID);
+		struct kvm_regs regs;
+
+		vcpu_run(vm, VCPU_ID);
+		TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
+			    "Got exit_reason other than KVM_EXIT_IO: %u (%s),\n",
+			    run->exit_reason,
+			    exit_reason_str(run->exit_reason));
+
+		vcpu_regs_get(vm, VCPU_ID, &regs);
+
+		switch (run->io.port) {
+		case PORT_ABORT:
+			TEST_ASSERT(false, "%s", (const char *) regs.rdi);
+			/* NOT REACHED */
+		case PORT_REPORT:
+			report(regs.rdi);
+			break;
+		case PORT_DONE:
+			goto done;
+		default:
+			TEST_ASSERT(false, "Unknown port 0x%x.", run->io.port);
+		}
+	}
+
+	kvm_vm_free(vm);
+done:
+	return 0;
+}