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

Pull kvm fixes from Paolo Bonzini:
 "Fixes for ARM and x86, plus selftest patches and nicer structs for
  nested state save/restore"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: nVMX: reorganize initial steps of vmx_set_nested_state
  KVM: arm/arm64: Fix emulated ptimer irq injection
  tests: kvm: Check for a kernel warning
  kvm: tests: Sort tests in the Makefile alphabetically
  KVM: x86/mmu: Allocate PAE root array when using SVM's 32-bit NPT
  KVM: x86: Modify struct kvm_nested_state to have explicit fields for data
  KVM: fix typo in documentation
  KVM: nVMX: use correct clean fields when copying from eVMCS
  KVM: arm/arm64: vgic: Fix kvm_device leak in vgic_its_destroy
  KVM: arm64: Filter out invalid core register IDs in KVM_GET_REG_LIST
  KVM: arm64: Implement vq_present() as a macro

Documentation/virtual/kvm/api.txt

@@ -1079,7 +1079,7 @@ yet and must be cleared on entry.
 
 4.35 KVM_SET_USER_MEMORY_REGION
 
-Capability: KVM_CAP_USER_MEM
+Capability: KVM_CAP_USER_MEMORY
 Architectures: all
 Type: vm ioctl
 Parameters: struct kvm_userspace_memory_region (in)
@@ -3857,43 +3857,59 @@ Type: vcpu ioctl
 Parameters: struct kvm_nested_state (in/out)
 Returns: 0 on success, -1 on error
 Errors:
-  E2BIG:     the total state size (including the fixed-size part of struct
-             kvm_nested_state) exceeds the value of 'size' specified by
+  E2BIG:     the total state size exceeds the value of 'size' specified by
              the user; the size required will be written into size.
 
 struct kvm_nested_state {
 	__u16 flags;
 	__u16 format;
 	__u32 size;
+
 	union {
-		struct kvm_vmx_nested_state vmx;
-		struct kvm_svm_nested_state svm;
+		struct kvm_vmx_nested_state_hdr vmx;
+		struct kvm_svm_nested_state_hdr svm;
+
+		/* Pad the header to 128 bytes.  */
 		__u8 pad[120];
-	};
-	__u8 data[0];
+	} hdr;
+
+	union {
+		struct kvm_vmx_nested_state_data vmx[0];
+		struct kvm_svm_nested_state_data svm[0];
+	} data;
 };
 
 #define KVM_STATE_NESTED_GUEST_MODE	0x00000001
 #define KVM_STATE_NESTED_RUN_PENDING	0x00000002
+#define KVM_STATE_NESTED_EVMCS		0x00000004
 
-#define KVM_STATE_NESTED_SMM_GUEST_MODE	0x00000001
-#define KVM_STATE_NESTED_SMM_VMXON	0x00000002
+#define KVM_STATE_NESTED_FORMAT_VMX		0
+#define KVM_STATE_NESTED_FORMAT_SVM		1
 
-struct kvm_vmx_nested_state {
+#define KVM_STATE_NESTED_VMX_VMCS_SIZE		0x1000
+
+#define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE	0x00000001
+#define KVM_STATE_NESTED_VMX_SMM_VMXON		0x00000002
+
+struct kvm_vmx_nested_state_hdr {
 	__u64 vmxon_pa;
-	__u64 vmcs_pa;
+	__u64 vmcs12_pa;
 
 	struct {
 		__u16 flags;
 	} smm;
 };
 
+struct kvm_vmx_nested_state_data {
+	__u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
+	__u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
+};
+
 This ioctl copies the vcpu's nested virtualization state from the kernel to
 userspace.
 
-The maximum size of the state, including the fixed-size part of struct
-kvm_nested_state, can be retrieved by passing KVM_CAP_NESTED_STATE to
-the KVM_CHECK_EXTENSION ioctl().
+The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE
+to the KVM_CHECK_EXTENSION ioctl().
 
 4.115 KVM_SET_NESTED_STATE
 
@@ -3903,8 +3919,8 @@ Type: vcpu ioctl
 Parameters: struct kvm_nested_state (in)
 Returns: 0 on success, -1 on error
 
-This copies the vcpu's kvm_nested_state struct from userspace to the kernel.  For
-the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE.
+This copies the vcpu's kvm_nested_state struct from userspace to the kernel.
+For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE.
 
 4.116 KVM_(UN)REGISTER_COALESCED_MMIO
 

arch/arm64/kvm/guest.c

@@ -70,10 +70,8 @@ static u64 core_reg_offset_from_id(u64 id)
 	return id & ~(KVM_REG_ARCH_MASK | KVM_REG_SIZE_MASK | KVM_REG_ARM_CORE);
 }
 
-static int validate_core_offset(const struct kvm_vcpu *vcpu,
-				const struct kvm_one_reg *reg)
+static int core_reg_size_from_offset(const struct kvm_vcpu *vcpu, u64 off)
 {
-	u64 off = core_reg_offset_from_id(reg->id);
 	int size;
 
 	switch (off) {
@@ -103,8 +101,7 @@ static int validate_core_offset(const struct kvm_vcpu *vcpu,
 		return -EINVAL;
 	}
 
-	if (KVM_REG_SIZE(reg->id) != size ||
-	    !IS_ALIGNED(off, size / sizeof(__u32)))
+	if (!IS_ALIGNED(off, size / sizeof(__u32)))
 		return -EINVAL;
 
 	/*
@@ -115,6 +112,21 @@ static int validate_core_offset(const struct kvm_vcpu *vcpu,
 	if (vcpu_has_sve(vcpu) && core_reg_offset_is_vreg(off))
 		return -EINVAL;
 
+	return size;
+}
+
+static int validate_core_offset(const struct kvm_vcpu *vcpu,
+				const struct kvm_one_reg *reg)
+{
+	u64 off = core_reg_offset_from_id(reg->id);
+	int size = core_reg_size_from_offset(vcpu, off);
+
+	if (size < 0)
+		return -EINVAL;
+
+	if (KVM_REG_SIZE(reg->id) != size)
+		return -EINVAL;
+
 	return 0;
 }
 
@@ -207,13 +219,7 @@ static int set_core_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)
 
 #define vq_word(vq) (((vq) - SVE_VQ_MIN) / 64)
 #define vq_mask(vq) ((u64)1 << ((vq) - SVE_VQ_MIN) % 64)
-
-static bool vq_present(
-	const u64 (*const vqs)[KVM_ARM64_SVE_VLS_WORDS],
-	unsigned int vq)
-{
-	return (*vqs)[vq_word(vq)] & vq_mask(vq);
-}
+#define vq_present(vqs, vq) ((vqs)[vq_word(vq)] & vq_mask(vq))
 
 static int get_sve_vls(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)
 {
@@ -258,7 +264,7 @@ static int set_sve_vls(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)
 
 	max_vq = 0;
 	for (vq = SVE_VQ_MIN; vq <= SVE_VQ_MAX; ++vq)
-		if (vq_present(&vqs, vq))
+		if (vq_present(vqs, vq))
 			max_vq = vq;
 
 	if (max_vq > sve_vq_from_vl(kvm_sve_max_vl))
@@ -272,7 +278,7 @@ static int set_sve_vls(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg)
 	 * maximum:
 	 */
 	for (vq = SVE_VQ_MIN; vq <= max_vq; ++vq)
-		if (vq_present(&vqs, vq) != sve_vq_available(vq))
+		if (vq_present(vqs, vq) != sve_vq_available(vq))
 			return -EINVAL;
 
 	/* Can't run with no vector lengths at all: */
@@ -453,19 +459,34 @@ static int copy_core_reg_indices(const struct kvm_vcpu *vcpu,
 {
 	unsigned int i;
 	int n = 0;
-	const u64 core_reg = KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE;
 
 	for (i = 0; i < sizeof(struct kvm_regs) / sizeof(__u32); i++) {
-		/*
-		 * The KVM_REG_ARM64_SVE regs must be used instead of
-		 * KVM_REG_ARM_CORE for accessing the FPSIMD V-registers on
-		 * SVE-enabled vcpus:
-		 */
-		if (vcpu_has_sve(vcpu) && core_reg_offset_is_vreg(i))
+		u64 reg = KVM_REG_ARM64 | KVM_REG_ARM_CORE | i;
+		int size = core_reg_size_from_offset(vcpu, i);
+
+		if (size < 0)
+			continue;
+
+		switch (size) {
+		case sizeof(__u32):
+			reg |= KVM_REG_SIZE_U32;
+			break;
+
+		case sizeof(__u64):
+			reg |= KVM_REG_SIZE_U64;
+			break;
+
+		case sizeof(__uint128_t):
+			reg |= KVM_REG_SIZE_U128;
+			break;
+
+		default:
+			WARN_ON(1);
 			continue;
+		}
 
 		if (uindices) {
-			if (put_user(core_reg | i, uindices))
+			if (put_user(reg, uindices))
 				return -EFAULT;
 			uindices++;
 		}

arch/x86/include/uapi/asm/kvm.h

@@ -383,6 +383,9 @@ struct kvm_sync_regs {
 #define KVM_X86_QUIRK_LAPIC_MMIO_HOLE	(1 << 2)
 #define KVM_X86_QUIRK_OUT_7E_INC_RIP	(1 << 3)
 
+#define KVM_STATE_NESTED_FORMAT_VMX	0
+#define KVM_STATE_NESTED_FORMAT_SVM	1	/* unused */
+
 #define KVM_STATE_NESTED_GUEST_MODE	0x00000001
 #define KVM_STATE_NESTED_RUN_PENDING	0x00000002
 #define KVM_STATE_NESTED_EVMCS		0x00000004
@@ -390,9 +393,16 @@ struct kvm_sync_regs {
 #define KVM_STATE_NESTED_SMM_GUEST_MODE	0x00000001
 #define KVM_STATE_NESTED_SMM_VMXON	0x00000002
 
-struct kvm_vmx_nested_state {
+#define KVM_STATE_NESTED_VMX_VMCS_SIZE	0x1000
+
+struct kvm_vmx_nested_state_data {
+	__u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
+	__u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
+};
+
+struct kvm_vmx_nested_state_hdr {
 	__u64 vmxon_pa;
-	__u64 vmcs_pa;
+	__u64 vmcs12_pa;
 
 	struct {
 		__u16 flags;
@@ -401,24 +411,25 @@ struct kvm_vmx_nested_state {
 
 /* for KVM_CAP_NESTED_STATE */
 struct kvm_nested_state {
-	/* KVM_STATE_* flags */
 	__u16 flags;
-
-	/* 0 for VMX, 1 for SVM.  */
 	__u16 format;
-
-	/* 128 for SVM, 128 + VMCS size for VMX.  */
 	__u32 size;
 
 	union {
-		/* VMXON, VMCS */
-		struct kvm_vmx_nested_state vmx;
+		struct kvm_vmx_nested_state_hdr vmx;
 
 		/* Pad the header to 128 bytes.  */
 		__u8 pad[120];
-	};
+	} hdr;
 
-	__u8 data[0];
+	/*
+	 * Define data region as 0 bytes to preserve backwards-compatability
+	 * to old definition of kvm_nested_state in order to avoid changing
+	 * KVM_{GET,PUT}_NESTED_STATE ioctl values.
+	 */
+	union {
+		struct kvm_vmx_nested_state_data vmx[0];
+	} data;
 };
 
 #endif /* _ASM_X86_KVM_H */

arch/x86/kvm/mmu.c

@@ -5602,14 +5602,18 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
 	struct page *page;
 	int i;
 
-	if (tdp_enabled)
-		return 0;
-
 	/*
-	 * When emulating 32-bit mode, cr3 is only 32 bits even on x86_64.
-	 * Therefore we need to allocate shadow page tables in the first
-	 * 4GB of memory, which happens to fit the DMA32 zone.
+	 * When using PAE paging, the four PDPTEs are treated as 'root' pages,
+	 * while the PDP table is a per-vCPU construct that's allocated at MMU
+	 * creation.  When emulating 32-bit mode, cr3 is only 32 bits even on
+	 * x86_64.  Therefore we need to allocate the PDP table in the first
+	 * 4GB of memory, which happens to fit the DMA32 zone.  Except for
+	 * SVM's 32-bit NPT support, TDP paging doesn't use PAE paging and can
+	 * skip allocating the PDP table.
 	 */
+	if (tdp_enabled && kvm_x86_ops->get_tdp_level(vcpu) > PT32E_ROOT_LEVEL)
+		return 0;
+
 	page = alloc_page(GFP_KERNEL_ACCOUNT | __GFP_DMA32);
 	if (!page)
 		return -ENOMEM;

arch/x86/kvm/vmx/nested.c

@@ -1397,7 +1397,7 @@ static int copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx)
 	}
 
 	if (unlikely(!(evmcs->hv_clean_fields &
-		       HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_PROC))) {
+		       HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_EXCPN))) {
 		vmcs12->exception_bitmap = evmcs->exception_bitmap;
 	}
 
@@ -1437,7 +1437,7 @@ static int copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx)
 	}
 
 	if (unlikely(!(evmcs->hv_clean_fields &
-		       HV_VMX_ENLIGHTENED_CLEAN_FIELD_HOST_GRP1))) {
+		       HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_GRP1))) {
 		vmcs12->pin_based_vm_exec_control =
 			evmcs->pin_based_vm_exec_control;
 		vmcs12->vm_exit_controls = evmcs->vm_exit_controls;
@@ -5226,14 +5226,16 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 	struct vmcs12 *vmcs12;
 	struct kvm_nested_state kvm_state = {
 		.flags = 0,
-		.format = 0,
+		.format = KVM_STATE_NESTED_FORMAT_VMX,
 		.size = sizeof(kvm_state),
-		.vmx.vmxon_pa = -1ull,
-		.vmx.vmcs_pa = -1ull,
+		.hdr.vmx.vmxon_pa = -1ull,
+		.hdr.vmx.vmcs12_pa = -1ull,
 	};
+	struct kvm_vmx_nested_state_data __user *user_vmx_nested_state =
+		&user_kvm_nested_state->data.vmx[0];
 
 	if (!vcpu)
-		return kvm_state.size + 2 * VMCS12_SIZE;
+		return kvm_state.size + sizeof(*user_vmx_nested_state);
 
 	vmx = to_vmx(vcpu);
 	vmcs12 = get_vmcs12(vcpu);
@@ -5243,23 +5245,23 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 
 	if (nested_vmx_allowed(vcpu) &&
 	    (vmx->nested.vmxon || vmx->nested.smm.vmxon)) {
-		kvm_state.vmx.vmxon_pa = vmx->nested.vmxon_ptr;
-		kvm_state.vmx.vmcs_pa = vmx->nested.current_vmptr;
+		kvm_state.hdr.vmx.vmxon_pa = vmx->nested.vmxon_ptr;
+		kvm_state.hdr.vmx.vmcs12_pa = vmx->nested.current_vmptr;
 
 		if (vmx_has_valid_vmcs12(vcpu)) {
-			kvm_state.size += VMCS12_SIZE;
+			kvm_state.size += sizeof(user_vmx_nested_state->vmcs12);
 
 			if (is_guest_mode(vcpu) &&
 			    nested_cpu_has_shadow_vmcs(vmcs12) &&
 			    vmcs12->vmcs_link_pointer != -1ull)
-				kvm_state.size += VMCS12_SIZE;
+				kvm_state.size += sizeof(user_vmx_nested_state->shadow_vmcs12);
 		}
 
 		if (vmx->nested.smm.vmxon)
-			kvm_state.vmx.smm.flags |= KVM_STATE_NESTED_SMM_VMXON;
+			kvm_state.hdr.vmx.smm.flags |= KVM_STATE_NESTED_SMM_VMXON;
 
 		if (vmx->nested.smm.guest_mode)
-			kvm_state.vmx.smm.flags |= KVM_STATE_NESTED_SMM_GUEST_MODE;
+			kvm_state.hdr.vmx.smm.flags |= KVM_STATE_NESTED_SMM_GUEST_MODE;
 
 		if (is_guest_mode(vcpu)) {
 			kvm_state.flags |= KVM_STATE_NESTED_GUEST_MODE;
@@ -5294,16 +5296,19 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 			copy_shadow_to_vmcs12(vmx);
 	}
 
+	BUILD_BUG_ON(sizeof(user_vmx_nested_state->vmcs12) < VMCS12_SIZE);
+	BUILD_BUG_ON(sizeof(user_vmx_nested_state->shadow_vmcs12) < VMCS12_SIZE);
+
 	/*
 	 * Copy over the full allocated size of vmcs12 rather than just the size
 	 * of the struct.
 	 */
-	if (copy_to_user(user_kvm_nested_state->data, vmcs12, VMCS12_SIZE))
+	if (copy_to_user(user_vmx_nested_state->vmcs12, vmcs12, VMCS12_SIZE))
 		return -EFAULT;
 
 	if (nested_cpu_has_shadow_vmcs(vmcs12) &&
 	    vmcs12->vmcs_link_pointer != -1ull) {
-		if (copy_to_user(user_kvm_nested_state->data + VMCS12_SIZE,
+		if (copy_to_user(user_vmx_nested_state->shadow_vmcs12,
 				 get_shadow_vmcs12(vcpu), VMCS12_SIZE))
 			return -EFAULT;
 	}
@@ -5331,33 +5336,35 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
 	struct vmcs12 *vmcs12;
 	u32 exit_qual;
+	struct kvm_vmx_nested_state_data __user *user_vmx_nested_state =
+		&user_kvm_nested_state->data.vmx[0];
 	int ret;
 
-	if (kvm_state->format != 0)
+	if (kvm_state->format != KVM_STATE_NESTED_FORMAT_VMX)
 		return -EINVAL;
 
-	if (!nested_vmx_allowed(vcpu))
-		return kvm_state->vmx.vmxon_pa == -1ull ? 0 : -EINVAL;
-
-	if (kvm_state->vmx.vmxon_pa == -1ull) {
-		if (kvm_state->vmx.smm.flags)
+	if (kvm_state->hdr.vmx.vmxon_pa == -1ull) {
+		if (kvm_state->hdr.vmx.smm.flags)
 			return -EINVAL;
 
-		if (kvm_state->vmx.vmcs_pa != -1ull)
+		if (kvm_state->hdr.vmx.vmcs12_pa != -1ull)
 			return -EINVAL;
 
-		vmx_leave_nested(vcpu);
-		return 0;
-	}
+		if (kvm_state->flags & ~KVM_STATE_NESTED_EVMCS)
+			return -EINVAL;
+	} else {
+		if (!nested_vmx_allowed(vcpu))
+			return -EINVAL;
 
-	if (!page_address_valid(vcpu, kvm_state->vmx.vmxon_pa))
-		return -EINVAL;
+		if (!page_address_valid(vcpu, kvm_state->hdr.vmx.vmxon_pa))
+			return -EINVAL;
+    	}
 
-	if ((kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
+	if ((kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
 	    (kvm_state->flags & KVM_STATE_NESTED_GUEST_MODE))
 		return -EINVAL;
 
-	if (kvm_state->vmx.smm.flags &
+	if (kvm_state->hdr.vmx.smm.flags &
 	    ~(KVM_STATE_NESTED_SMM_GUEST_MODE | KVM_STATE_NESTED_SMM_VMXON))
 		return -EINVAL;
 
@@ -5366,21 +5373,25 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	 * nor can VMLAUNCH/VMRESUME be pending.  Outside SMM, SMM flags
 	 * must be zero.
 	 */
-	if (is_smm(vcpu) ? kvm_state->flags : kvm_state->vmx.smm.flags)
+	if (is_smm(vcpu) ? kvm_state->flags : kvm_state->hdr.vmx.smm.flags)
 		return -EINVAL;
 
-	if ((kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
-	    !(kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON))
+	if ((kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
+	    !(kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON))
 		return -EINVAL;
 
 	vmx_leave_nested(vcpu);
-	if (kvm_state->vmx.vmxon_pa == -1ull)
-		return 0;
+	if (kvm_state->flags & KVM_STATE_NESTED_EVMCS) {
+		if (!nested_vmx_allowed(vcpu))
+			return -EINVAL;
 
-	if (kvm_state->flags & KVM_STATE_NESTED_EVMCS)
 		nested_enable_evmcs(vcpu, NULL);
+	}
+
+	if (kvm_state->hdr.vmx.vmxon_pa == -1ull)
+		return 0;
 
-	vmx->nested.vmxon_ptr = kvm_state->vmx.vmxon_pa;
+	vmx->nested.vmxon_ptr = kvm_state->hdr.vmx.vmxon_pa;
 	ret = enter_vmx_operation(vcpu);
 	if (ret)
 		return ret;
@@ -5389,12 +5400,12 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	if (kvm_state->size < sizeof(*kvm_state) + sizeof(*vmcs12))
 		return 0;
 
-	if (kvm_state->vmx.vmcs_pa != -1ull) {
-		if (kvm_state->vmx.vmcs_pa == kvm_state->vmx.vmxon_pa ||
-		    !page_address_valid(vcpu, kvm_state->vmx.vmcs_pa))
+	if (kvm_state->hdr.vmx.vmcs12_pa != -1ull) {
+		if (kvm_state->hdr.vmx.vmcs12_pa == kvm_state->hdr.vmx.vmxon_pa ||
+		    !page_address_valid(vcpu, kvm_state->hdr.vmx.vmcs12_pa))
 			return -EINVAL;
 
-		set_current_vmptr(vmx, kvm_state->vmx.vmcs_pa);
+		set_current_vmptr(vmx, kvm_state->hdr.vmx.vmcs12_pa);
 	} else if (kvm_state->flags & KVM_STATE_NESTED_EVMCS) {
 		/*
 		 * Sync eVMCS upon entry as we may not have
@@ -5405,16 +5416,16 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 		return -EINVAL;
 	}
 
-	if (kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON) {
+	if (kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON) {
 		vmx->nested.smm.vmxon = true;
 		vmx->nested.vmxon = false;
 
-		if (kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE)
+		if (kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE)
 			vmx->nested.smm.guest_mode = true;
 	}
 
 	vmcs12 = get_vmcs12(vcpu);
-	if (copy_from_user(vmcs12, user_kvm_nested_state->data, sizeof(*vmcs12)))
+	if (copy_from_user(vmcs12, user_vmx_nested_state->vmcs12, sizeof(*vmcs12)))
 		return -EFAULT;
 
 	if (vmcs12->hdr.revision_id != VMCS12_REVISION)
@@ -5431,12 +5442,14 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	    vmcs12->vmcs_link_pointer != -1ull) {
 		struct vmcs12 *shadow_vmcs12 = get_shadow_vmcs12(vcpu);
 
-		if (kvm_state->size < sizeof(*kvm_state) + VMCS12_SIZE + sizeof(*vmcs12))
+		if (kvm_state->size <
+		    sizeof(*kvm_state) +
+		    sizeof(user_vmx_nested_state->vmcs12) + sizeof(*shadow_vmcs12))
 			goto error_guest_mode;
 
 		if (copy_from_user(shadow_vmcs12,
-				   user_kvm_nested_state->data + VMCS12_SIZE,
-				   sizeof(*vmcs12))) {
+				   user_vmx_nested_state->shadow_vmcs12,
+				   sizeof(*shadow_vmcs12))) {
 			ret = -EFAULT;
 			goto error_guest_mode;
 		}

arch/x86/kvm/vmx/vmcs12.h

@@ -201,9 +201,10 @@ struct __packed vmcs12 {
 /*
  * VMCS12_SIZE is the number of bytes L1 should allocate for the VMXON region
  * and any VMCS region. Although only sizeof(struct vmcs12) are used by the
- * current implementation, 4K are reserved to avoid future complications.
+ * current implementation, 4K are reserved to avoid future complications and
+ * to preserve userspace ABI.
  */
-#define VMCS12_SIZE 0x1000
+#define VMCS12_SIZE		KVM_STATE_NESTED_VMX_VMCS_SIZE
 
 /*
  * VMCS12_MAX_FIELD_INDEX is the highest index value used in any

tools/arch/x86/include/uapi/asm/kvm.h

@@ -392,7 +392,7 @@ struct kvm_sync_regs {
 
 struct kvm_vmx_nested_state {
 	__u64 vmxon_pa;
-	__u64 vmcs_pa;
+	__u64 vmcs12_pa;
 
 	struct {
 		__u16 flags;

tools/testing/selftests/kvm/.gitignore

@@ -2,6 +2,7 @@
 /x86_64/evmcs_test
 /x86_64/hyperv_cpuid
 /x86_64/kvm_create_max_vcpus
+/x86_64/mmio_warning_test
 /x86_64/platform_info_test
 /x86_64/set_sregs_test
 /x86_64/smm_test

tools/testing/selftests/kvm/Makefile

@@ -11,23 +11,24 @@ LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/ucall.c lib/sparsebi
 LIBKVM_x86_64 = lib/x86_64/processor.c lib/x86_64/vmx.c
 LIBKVM_aarch64 = lib/aarch64/processor.c
 
-TEST_GEN_PROGS_x86_64 = x86_64/platform_info_test
-TEST_GEN_PROGS_x86_64 += x86_64/set_sregs_test
-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/cr4_cpuid_sync_test
 TEST_GEN_PROGS_x86_64 += x86_64/evmcs_test
 TEST_GEN_PROGS_x86_64 += x86_64/hyperv_cpuid
-TEST_GEN_PROGS_x86_64 += x86_64/vmx_close_while_nested_test
-TEST_GEN_PROGS_x86_64 += x86_64/smm_test
 TEST_GEN_PROGS_x86_64 += x86_64/kvm_create_max_vcpus
+TEST_GEN_PROGS_x86_64 += x86_64/mmio_warning_test
+TEST_GEN_PROGS_x86_64 += x86_64/platform_info_test
+TEST_GEN_PROGS_x86_64 += x86_64/set_sregs_test
+TEST_GEN_PROGS_x86_64 += x86_64/smm_test
+TEST_GEN_PROGS_x86_64 += x86_64/state_test
+TEST_GEN_PROGS_x86_64 += x86_64/sync_regs_test
+TEST_GEN_PROGS_x86_64 += x86_64/vmx_close_while_nested_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_set_nested_state_test
-TEST_GEN_PROGS_x86_64 += dirty_log_test
+TEST_GEN_PROGS_x86_64 += x86_64/vmx_tsc_adjust_test
 TEST_GEN_PROGS_x86_64 += clear_dirty_log_test
+TEST_GEN_PROGS_x86_64 += dirty_log_test
 
-TEST_GEN_PROGS_aarch64 += dirty_log_test
 TEST_GEN_PROGS_aarch64 += clear_dirty_log_test
+TEST_GEN_PROGS_aarch64 += dirty_log_test
 
 TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(UNAME_M))
 LIBKVM += $(LIBKVM_$(UNAME_M))

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

@@ -139,6 +139,8 @@ struct kvm_vm *vm_create_default(uint32_t vcpuid, uint64_t extra_mem_size,
 				 void *guest_code);
 void vm_vcpu_add_default(struct kvm_vm *vm, uint32_t vcpuid, void *guest_code);
 
+bool vm_is_unrestricted_guest(struct kvm_vm *vm);
+
 struct kvm_userspace_memory_region *
 kvm_userspace_memory_region_find(struct kvm_vm *vm, uint64_t start,
 				 uint64_t end);

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

@@ -303,6 +303,8 @@ static inline unsigned long get_xmm(int n)
 	return 0;
 }
 
+bool is_intel_cpu(void);
+
 struct kvm_x86_state;
 struct kvm_x86_state *vcpu_save_state(struct kvm_vm *vm, uint32_t vcpuid);
 void vcpu_load_state(struct kvm_vm *vm, uint32_t vcpuid,

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

@@ -1583,3 +1583,39 @@ void *addr_gva2hva(struct kvm_vm *vm, vm_vaddr_t gva)
 {
 	return addr_gpa2hva(vm, addr_gva2gpa(vm, gva));
 }
+
+/*
+ * Is Unrestricted Guest
+ *
+ * Input Args:
+ *   vm - Virtual Machine
+ *
+ * Output Args: None
+ *
+ * Return: True if the unrestricted guest is set to 'Y', otherwise return false.
+ *
+ * Check if the unrestricted guest flag is enabled.
+ */
+bool vm_is_unrestricted_guest(struct kvm_vm *vm)
+{
+	char val = 'N';
+	size_t count;
+	FILE *f;
+
+	if (vm == NULL) {
+		/* Ensure that the KVM vendor-specific module is loaded. */
+		f = fopen(KVM_DEV_PATH, "r");
+		TEST_ASSERT(f != NULL, "Error in opening KVM dev file: %d",
+			    errno);
+		fclose(f);
+	}
+
+	f = fopen("/sys/module/kvm_intel/parameters/unrestricted_guest", "r");
+	if (f) {
+		count = fread(&val, sizeof(char), 1, f);
+		TEST_ASSERT(count == 1, "Unable to read from param file.");
+		fclose(f);
+	}
+
+	return val == 'Y';
+}

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

@@ -1137,3 +1137,19 @@ void vcpu_load_state(struct kvm_vm *vm, uint32_t vcpuid, struct kvm_x86_state *s
 			r);
 	}
 }
+
+bool is_intel_cpu(void)
+{
+	int eax, ebx, ecx, edx;
+	const uint32_t *chunk;
+	const int leaf = 0;
+
+	__asm__ __volatile__(
+		"cpuid"
+		: /* output */ "=a"(eax), "=b"(ebx),
+		  "=c"(ecx), "=d"(edx)
+		: /* input */ "0"(leaf), "2"(0));
+
+	chunk = (const uint32_t *)("GenuineIntel");
+	return (ebx == chunk[0] && edx == chunk[1] && ecx == chunk[2]);
+}

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

+/*
+ * mmio_warning_test
+ *
+ * Copyright (C) 2019, Google LLC.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ *
+ * Test that we don't get a kernel warning when we call KVM_RUN after a
+ * triple fault occurs.  To get the triple fault to occur we call KVM_RUN
+ * on a VCPU that hasn't been properly setup.
+ *
+ */
+
+#define _GNU_SOURCE
+#include <fcntl.h>
+#include <kvm_util.h>
+#include <linux/kvm.h>
+#include <processor.h>
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <test_util.h>
+#include <unistd.h>
+
+#define NTHREAD 4
+#define NPROCESS 5
+
+struct thread_context {
+	int kvmcpu;
+	struct kvm_run *run;
+};
+
+void *thr(void *arg)
+{
+	struct thread_context *tc = (struct thread_context *)arg;
+	int res;
+	int kvmcpu = tc->kvmcpu;
+	struct kvm_run *run = tc->run;
+
+	res = ioctl(kvmcpu, KVM_RUN, 0);
+	printf("ret1=%d exit_reason=%d suberror=%d\n",
+		res, run->exit_reason, run->internal.suberror);
+
+	return 0;
+}
+
+void test(void)
+{
+	int i, kvm, kvmvm, kvmcpu;
+	pthread_t th[NTHREAD];
+	struct kvm_run *run;
+	struct thread_context tc;
+
+	kvm = open("/dev/kvm", O_RDWR);
+	TEST_ASSERT(kvm != -1, "failed to open /dev/kvm");
+	kvmvm = ioctl(kvm, KVM_CREATE_VM, 0);
+	TEST_ASSERT(kvmvm != -1, "KVM_CREATE_VM failed");
+	kvmcpu = ioctl(kvmvm, KVM_CREATE_VCPU, 0);
+	TEST_ASSERT(kvmcpu != -1, "KVM_CREATE_VCPU failed");
+	run = (struct kvm_run *)mmap(0, 4096, PROT_READ|PROT_WRITE, MAP_SHARED,
+				    kvmcpu, 0);
+	tc.kvmcpu = kvmcpu;
+	tc.run = run;
+	srand(getpid());
+	for (i = 0; i < NTHREAD; i++) {
+		pthread_create(&th[i], NULL, thr, (void *)(uintptr_t)&tc);
+		usleep(rand() % 10000);
+	}
+	for (i = 0; i < NTHREAD; i++)
+		pthread_join(th[i], NULL);
+}
+
+int get_warnings_count(void)
+{
+	int warnings;
+	FILE *f;
+
+	f = popen("dmesg | grep \"WARNING:\" | wc -l", "r");
+	fscanf(f, "%d", &warnings);
+	fclose(f);
+
+	return warnings;
+}
+
+int main(void)
+{
+	int warnings_before, warnings_after;
+
+	if (!is_intel_cpu()) {
+		printf("Must be run on an Intel CPU, skipping test\n");
+		exit(KSFT_SKIP);
+	}
+
+	if (vm_is_unrestricted_guest(NULL)) {
+		printf("Unrestricted guest must be disabled, skipping test\n");
+		exit(KSFT_SKIP);
+	}
+
+	warnings_before = get_warnings_count();
+
+	for (int i = 0; i < NPROCESS; ++i) {
+		int status;
+		int pid = fork();
+
+		if (pid < 0)
+			exit(1);
+		if (pid == 0) {
+			test();
+			exit(0);
+		}
+		while (waitpid(pid, &status, __WALL) != pid)
+			;
+	}
+
+	warnings_after = get_warnings_count();
+	TEST_ASSERT(warnings_before == warnings_after,
+		   "Warnings found in kernel.  Run 'dmesg' to inspect them.");
+
+	return 0;
+}

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

@@ -75,7 +75,7 @@ void set_revision_id_for_vmcs12(struct kvm_nested_state *state,
 				u32 vmcs12_revision)
 {
 	/* Set revision_id in vmcs12 to vmcs12_revision. */
-	memcpy(state->data, &vmcs12_revision, sizeof(u32));
+	memcpy(&state->data, &vmcs12_revision, sizeof(u32));
 }
 
 void set_default_state(struct kvm_nested_state *state)
@@ -95,9 +95,9 @@ void set_default_vmx_state(struct kvm_nested_state *state, int size)
 			KVM_STATE_NESTED_EVMCS;
 	state->format = 0;
 	state->size = size;
-	state->vmx.vmxon_pa = 0x1000;
-	state->vmx.vmcs_pa = 0x2000;
-	state->vmx.smm.flags = 0;
+	state->hdr.vmx.vmxon_pa = 0x1000;
+	state->hdr.vmx.vmcs12_pa = 0x2000;
+	state->hdr.vmx.smm.flags = 0;
 	set_revision_id_for_vmcs12(state, VMCS12_REVISION);
 }
 
@@ -123,39 +123,47 @@ void test_vmx_nested_state(struct kvm_vm *vm)
 	/*
 	 * We cannot virtualize anything if the guest does not have VMX
 	 * enabled.  We expect KVM_SET_NESTED_STATE to return 0 if vmxon_pa
-	 * is set to -1ull.
+	 * is set to -1ull, but the flags must be zero.
 	 */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = -1ull;
+	state->hdr.vmx.vmxon_pa = -1ull;
+	test_nested_state_expect_einval(vm, state);
+
+	state->hdr.vmx.vmcs12_pa = -1ull;
+	state->flags = KVM_STATE_NESTED_EVMCS;
+	test_nested_state_expect_einval(vm, state);
+
+	state->flags = 0;
 	test_nested_state(vm, state);
 
 	/* Enable VMX in the guest CPUID. */
 	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
 
-	/* It is invalid to have vmxon_pa == -1ull and SMM flags non-zero. */
+	/*
+	 * Setting vmxon_pa == -1ull and vmcs_pa == -1ull exits early without
+	 * setting the nested state but flags other than eVMCS must be clear.
+	 */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = -1ull;
-	state->vmx.smm.flags = 1;
+	state->hdr.vmx.vmxon_pa = -1ull;
+	state->hdr.vmx.vmcs12_pa = -1ull;
 	test_nested_state_expect_einval(vm, state);
 
-	/* It is invalid to have vmxon_pa == -1ull and vmcs_pa != -1ull. */
-	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = -1ull;
-	state->vmx.vmcs_pa = 0;
+	state->flags = KVM_STATE_NESTED_EVMCS;
+	test_nested_state(vm, state);
+
+	/* It is invalid to have vmxon_pa == -1ull and SMM flags non-zero. */
+	state->hdr.vmx.smm.flags = 1;
 	test_nested_state_expect_einval(vm, state);
 
-	/*
-	 * Setting vmxon_pa == -1ull and vmcs_pa == -1ull exits early without
-	 * setting the nested state.
-	 */
+	/* It is invalid to have vmxon_pa == -1ull and vmcs_pa != -1ull. */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = -1ull;
-	state->vmx.vmcs_pa = -1ull;
-	test_nested_state(vm, state);
+	state->hdr.vmx.vmxon_pa = -1ull;
+	state->flags = 0;
+	test_nested_state_expect_einval(vm, state);
 
 	/* It is invalid to have vmxon_pa set to a non-page aligned address. */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = 1;
+	state->hdr.vmx.vmxon_pa = 1;
 	test_nested_state_expect_einval(vm, state);
 
 	/*
@@ -165,7 +173,7 @@ void test_vmx_nested_state(struct kvm_vm *vm)
 	set_default_vmx_state(state, state_sz);
 	state->flags = KVM_STATE_NESTED_GUEST_MODE  |
 		      KVM_STATE_NESTED_RUN_PENDING;
-	state->vmx.smm.flags = KVM_STATE_NESTED_SMM_GUEST_MODE;
+	state->hdr.vmx.smm.flags = KVM_STATE_NESTED_SMM_GUEST_MODE;
 	test_nested_state_expect_einval(vm, state);
 
 	/*
@@ -174,14 +182,14 @@ void test_vmx_nested_state(struct kvm_vm *vm)
 	 *	KVM_STATE_NESTED_SMM_VMXON
 	 */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.smm.flags = ~(KVM_STATE_NESTED_SMM_GUEST_MODE |
+	state->hdr.vmx.smm.flags = ~(KVM_STATE_NESTED_SMM_GUEST_MODE |
 				KVM_STATE_NESTED_SMM_VMXON);
 	test_nested_state_expect_einval(vm, state);
 
 	/* Outside SMM, SMM flags must be zero. */
 	set_default_vmx_state(state, state_sz);
 	state->flags = 0;
-	state->vmx.smm.flags = KVM_STATE_NESTED_SMM_GUEST_MODE;
+	state->hdr.vmx.smm.flags = KVM_STATE_NESTED_SMM_GUEST_MODE;
 	test_nested_state_expect_einval(vm, state);
 
 	/* Size must be large enough to fit kvm_nested_state and vmcs12. */
@@ -191,8 +199,8 @@ void test_vmx_nested_state(struct kvm_vm *vm)
 
 	/* vmxon_pa cannot be the same address as vmcs_pa. */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = 0;
-	state->vmx.vmcs_pa = 0;
+	state->hdr.vmx.vmxon_pa = 0;
+	state->hdr.vmx.vmcs12_pa = 0;
 	test_nested_state_expect_einval(vm, state);
 
 	/* The revision id for vmcs12 must be VMCS12_REVISION. */
@@ -205,16 +213,16 @@ void test_vmx_nested_state(struct kvm_vm *vm)
 	 * it again.
 	 */
 	set_default_vmx_state(state, state_sz);
-	state->vmx.vmxon_pa = -1ull;
-	state->vmx.vmcs_pa = -1ull;
+	state->hdr.vmx.vmxon_pa = -1ull;
+	state->hdr.vmx.vmcs12_pa = -1ull;
 	state->flags = 0;
 	test_nested_state(vm, state);
 	vcpu_nested_state_get(vm, VCPU_ID, state);
 	TEST_ASSERT(state->size >= sizeof(*state) && state->size <= state_sz,
 		    "Size must be between %d and %d.  The size returned was %d.",
 		    sizeof(*state), state_sz, state->size);
-	TEST_ASSERT(state->vmx.vmxon_pa == -1ull, "vmxon_pa must be -1ull.");
-	TEST_ASSERT(state->vmx.vmcs_pa == -1ull, "vmcs_pa must be -1ull.");
+	TEST_ASSERT(state->hdr.vmx.vmxon_pa == -1ull, "vmxon_pa must be -1ull.");
+	TEST_ASSERT(state->hdr.vmx.vmcs12_pa == -1ull, "vmcs_pa must be -1ull.");
 
 	free(state);
 }

virt/kvm/arm/arch_timer.c

@@ -309,14 +309,15 @@ static void kvm_timer_update_irq(struct kvm_vcpu *vcpu, bool new_level,
 	}
 }
 
+/* Only called for a fully emulated timer */
 static void timer_emulate(struct arch_timer_context *ctx)
 {
 	bool should_fire = kvm_timer_should_fire(ctx);
 
 	trace_kvm_timer_emulate(ctx, should_fire);
 
-	if (should_fire) {
-		kvm_timer_update_irq(ctx->vcpu, true, ctx);
+	if (should_fire != ctx->irq.level) {
+		kvm_timer_update_irq(ctx->vcpu, should_fire, ctx);
 		return;
 	}
 

virt/kvm/arm/vgic/vgic-its.c

@@ -1734,6 +1734,7 @@ static void vgic_its_destroy(struct kvm_device *kvm_dev)
 
 	mutex_unlock(&its->its_lock);
 	kfree(its);
+	kfree(kvm_dev);/* alloc by kvm_ioctl_create_device, free by .destroy */
 }
 
 static int vgic_its_has_attr_regs(struct kvm_device *dev,