diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index ce60f4c38843f..ffd45b68e1d46 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -639,7 +639,7 @@ struct kvm_vcpu_arch {
 	struct kvm_cpuid_entry2 cpuid_entries[KVM_MAX_CPUID_ENTRIES];
 
 	int maxphyaddr;
-	int tdp_level;
+	int max_tdp_level;
 
 	/* emulate context */
 
@@ -1133,7 +1133,7 @@ struct kvm_x86_ops {
 	int (*sync_pir_to_irr)(struct kvm_vcpu *vcpu);
 	int (*set_tss_addr)(struct kvm *kvm, unsigned int addr);
 	int (*set_identity_map_addr)(struct kvm *kvm, u64 ident_addr);
-	int (*get_tdp_level)(struct kvm_vcpu *vcpu);
+	int (*get_max_tdp_level)(void);
 	u64 (*get_mt_mask)(struct kvm_vcpu *vcpu, gfn_t gfn, bool is_mmio);
 
 	void (*load_mmu_pgd)(struct kvm_vcpu *vcpu, unsigned long pgd,
diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
index 7d92854082a14..fa873e3e6e90e 100644
--- a/arch/x86/kvm/cpuid.c
+++ b/arch/x86/kvm/cpuid.c
@@ -140,9 +140,7 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
 		vcpu->arch.guest_supported_xcr0 =
 			(best->eax | ((u64)best->edx << 32)) & supported_xcr0;
 
-	/* Note, maxphyaddr must be updated before tdp_level. */
 	vcpu->arch.maxphyaddr = cpuid_query_maxphyaddr(vcpu);
-	vcpu->arch.tdp_level = kvm_x86_ops.get_tdp_level(vcpu);
 	kvm_mmu_reset_context(vcpu);
 
 	kvm_pmu_refresh(vcpu);
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 0fb033ce6cc57..559b4b92b5e27 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4846,13 +4846,22 @@ static union kvm_mmu_role kvm_calc_mmu_role_common(struct kvm_vcpu *vcpu,
 	return role;
 }
 
+static inline int kvm_mmu_get_tdp_level(struct kvm_vcpu *vcpu)
+{
+	/* Use 5-level TDP if and only if it's useful/necessary. */
+	if (vcpu->arch.max_tdp_level == 5 && cpuid_maxphyaddr(vcpu) <= 48)
+		return 4;
+
+	return vcpu->arch.max_tdp_level;
+}
+
 static union kvm_mmu_role
 kvm_calc_tdp_mmu_root_page_role(struct kvm_vcpu *vcpu, bool base_only)
 {
 	union kvm_mmu_role role = kvm_calc_mmu_role_common(vcpu, base_only);
 
 	role.base.ad_disabled = (shadow_accessed_mask == 0);
-	role.base.level = vcpu->arch.tdp_level;
+	role.base.level = kvm_mmu_get_tdp_level(vcpu);
 	role.base.direct = true;
 	role.base.gpte_is_8_bytes = true;
 
@@ -4873,7 +4882,7 @@ static void init_kvm_tdp_mmu(struct kvm_vcpu *vcpu)
 	context->sync_page = nonpaging_sync_page;
 	context->invlpg = NULL;
 	context->update_pte = nonpaging_update_pte;
-	context->shadow_root_level = vcpu->arch.tdp_level;
+	context->shadow_root_level = kvm_mmu_get_tdp_level(vcpu);
 	context->direct_map = true;
 	context->get_guest_pgd = get_cr3;
 	context->get_pdptr = kvm_pdptr_read;
@@ -4973,7 +4982,7 @@ kvm_calc_shadow_npt_root_page_role(struct kvm_vcpu *vcpu)
 		kvm_calc_shadow_root_page_role_common(vcpu, false);
 
 	role.base.direct = false;
-	role.base.level = vcpu->arch.tdp_level;
+	role.base.level = kvm_mmu_get_tdp_level(vcpu);
 
 	return role;
 }
@@ -5683,7 +5692,7 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)
 	 * SVM's 32-bit NPT support, TDP paging doesn't use PAE paging and can
 	 * skip allocating the PDP table.
 	 */
-	if (tdp_enabled && vcpu->arch.tdp_level > PT32E_ROOT_LEVEL)
+	if (tdp_enabled && kvm_mmu_get_tdp_level(vcpu) > PT32E_ROOT_LEVEL)
 		return 0;
 
 	page = alloc_page(GFP_KERNEL_ACCOUNT | __GFP_DMA32);
diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index c70d7dd333061..c94faca46e760 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -254,7 +254,7 @@ static inline void invlpga(unsigned long addr, u32 asid)
 	asm volatile (__ex("invlpga %1, %0") : : "c"(asid), "a"(addr));
 }
 
-static int get_npt_level(struct kvm_vcpu *vcpu)
+static int get_max_npt_level(void)
 {
 #ifdef CONFIG_X86_64
 	return PT64_ROOT_4LEVEL;
@@ -4109,7 +4109,7 @@ static struct kvm_x86_ops svm_x86_ops __initdata = {
 
 	.set_tss_addr = svm_set_tss_addr,
 	.set_identity_map_addr = svm_set_identity_map_addr,
-	.get_tdp_level = get_npt_level,
+	.get_max_tdp_level = get_max_npt_level,
 	.get_mt_mask = svm_get_mt_mask,
 
 	.get_exit_info = svm_get_exit_info,
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index da75878171cea..c0b1c7bd1248a 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -3065,9 +3065,9 @@ void vmx_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
 	vmx->emulation_required = emulation_required(vcpu);
 }
 
-static int vmx_get_tdp_level(struct kvm_vcpu *vcpu)
+static int vmx_get_max_tdp_level(void)
 {
-	if (cpu_has_vmx_ept_5levels() && (cpuid_maxphyaddr(vcpu) > 48))
+	if (cpu_has_vmx_ept_5levels())
 		return 5;
 	return 4;
 }
@@ -7959,7 +7959,7 @@ static struct kvm_x86_ops vmx_x86_ops __initdata = {
 
 	.set_tss_addr = vmx_set_tss_addr,
 	.set_identity_map_addr = vmx_set_identity_map_addr,
-	.get_tdp_level = vmx_get_tdp_level,
+	.get_max_tdp_level = vmx_get_max_tdp_level,
 	.get_mt_mask = vmx_get_mt_mask,
 
 	.get_exit_info = vmx_get_exit_info,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 95ef629228691..41f43bb716c1b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -9520,7 +9520,7 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu)
 	fx_init(vcpu);
 
 	vcpu->arch.maxphyaddr = cpuid_query_maxphyaddr(vcpu);
-	vcpu->arch.tdp_level = kvm_x86_ops.get_tdp_level(vcpu);
+	vcpu->arch.max_tdp_level = kvm_x86_ops.get_max_tdp_level();
 
 	vcpu->arch.pat = MSR_IA32_CR_PAT_DEFAULT;