diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
index d4e4212ff596..d7cf66573aca 100644
--- a/arch/arm64/include/asm/kvm_emulate.h
+++ b/arch/arm64/include/asm/kvm_emulate.h
@@ -662,4 +662,28 @@ static inline bool guest_hyp_sve_traps_enabled(const struct kvm_vcpu *vcpu)
 {
 	return __guest_hyp_cptr_xen_trap_enabled(vcpu, ZEN);
 }
+
+static inline void vcpu_set_hcrx(struct kvm_vcpu *vcpu)
+{
+	struct kvm *kvm = vcpu->kvm;
+
+	if (cpus_have_final_cap(ARM64_HAS_HCX)) {
+		/*
+		 * In general, all HCRX_EL2 bits are gated by a feature.
+		 * The only reason we can set SMPME without checking any
+		 * feature is that its effects are not directly observable
+		 * from the guest.
+		 */
+		vcpu->arch.hcrx_el2 = HCRX_EL2_SMPME;
+
+		if (kvm_has_feat(kvm, ID_AA64ISAR2_EL1, MOPS, IMP))
+			vcpu->arch.hcrx_el2 |= (HCRX_EL2_MSCEn | HCRX_EL2_MCE2);
+
+		if (kvm_has_tcr2(kvm))
+			vcpu->arch.hcrx_el2 |= HCRX_EL2_TCR2En;
+
+		if (kvm_has_fpmr(kvm))
+			vcpu->arch.hcrx_el2 |= HCRX_EL2_EnFPM;
+	}
+}
 #endif /* __ARM64_KVM_EMULATE_H__ */
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index f49b680e032d..a73fe3671e3d 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -87,6 +87,9 @@ struct kvm_hyp_memcache {
 	phys_addr_t head;
 	unsigned long nr_pages;
 	struct pkvm_mapping *mapping; /* only used from EL1 */
+
+#define	HYP_MEMCACHE_ACCOUNT_STAGE2	BIT(1)
+	unsigned long flags;
 };
 
 static inline void push_hyp_memcache(struct kvm_hyp_memcache *mc,
@@ -247,6 +250,7 @@ typedef unsigned int pkvm_handle_t;
 struct kvm_protected_vm {
 	pkvm_handle_t handle;
 	struct kvm_hyp_memcache teardown_mc;
+	struct kvm_hyp_memcache stage2_teardown_mc;
 	bool enabled;
 };
 
@@ -902,6 +906,8 @@ struct kvm_vcpu_arch {
 #define VCPU_INITIALIZED	__vcpu_single_flag(cflags, BIT(0))
 /* SVE config completed */
 #define VCPU_SVE_FINALIZED	__vcpu_single_flag(cflags, BIT(1))
+/* pKVM VCPU setup completed */
+#define VCPU_PKVM_FINALIZED	__vcpu_single_flag(cflags, BIT(2))
 
 /* Exception pending */
 #define PENDING_EXCEPTION	__vcpu_single_flag(iflags, BIT(0))
diff --git a/arch/arm64/include/asm/kvm_pkvm.h b/arch/arm64/include/asm/kvm_pkvm.h
index eb65f12e81d9..abd693ce5b93 100644
--- a/arch/arm64/include/asm/kvm_pkvm.h
+++ b/arch/arm64/include/asm/kvm_pkvm.h
@@ -19,6 +19,7 @@
 int pkvm_init_host_vm(struct kvm *kvm);
 int pkvm_create_hyp_vm(struct kvm *kvm);
 void pkvm_destroy_hyp_vm(struct kvm *kvm);
+int pkvm_create_hyp_vcpu(struct kvm_vcpu *vcpu);
 
 /*
  * This functions as an allow-list of protected VM capabilities.
diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index d466b8680bb7..b7192a23218b 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -866,6 +866,10 @@ int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
 		ret = pkvm_create_hyp_vm(kvm);
 		if (ret)
 			return ret;
+
+		ret = pkvm_create_hyp_vcpu(vcpu);
+		if (ret)
+			return ret;
 	}
 
 	mutex_lock(&kvm->arch.config_lock);
diff --git a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
index 978f38c386ee..ea0a704da9b8 100644
--- a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
+++ b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
@@ -56,7 +56,7 @@ void handle_host_mem_abort(struct kvm_cpu_context *host_ctxt);
 
 int hyp_pin_shared_mem(void *from, void *to);
 void hyp_unpin_shared_mem(void *from, void *to);
-void reclaim_guest_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc);
+void reclaim_pgtable_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc);
 int refill_memcache(struct kvm_hyp_memcache *mc, unsigned long min_pages,
 		    struct kvm_hyp_memcache *host_mc);
 
diff --git a/arch/arm64/kvm/hyp/include/nvhe/pkvm.h b/arch/arm64/kvm/hyp/include/nvhe/pkvm.h
index e42bf68c8848..ce31d3b73603 100644
--- a/arch/arm64/kvm/hyp/include/nvhe/pkvm.h
+++ b/arch/arm64/kvm/hyp/include/nvhe/pkvm.h
@@ -43,12 +43,6 @@ struct pkvm_hyp_vm {
 	struct hyp_pool pool;
 	hyp_spinlock_t lock;
 
-	/*
-	 * The number of vcpus initialized and ready to run.
-	 * Modifying this is protected by 'vm_table_lock'.
-	 */
-	unsigned int nr_vcpus;
-
 	/* Array of the hyp vCPU structures for this VM. */
 	struct pkvm_hyp_vcpu *vcpus[];
 };
diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
index 19c3c631708c..f34f11c720d7 100644
--- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
+++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
@@ -266,7 +266,7 @@ int kvm_guest_prepare_stage2(struct pkvm_hyp_vm *vm, void *pgd)
 	return 0;
 }
 
-void reclaim_guest_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc)
+void reclaim_pgtable_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc)
 {
 	struct hyp_page *page;
 	void *addr;
diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c
index 64a4a4c95bc3..5a335a51deca 100644
--- a/arch/arm64/kvm/hyp/nvhe/pkvm.c
+++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c
@@ -167,8 +167,13 @@ static int pkvm_vcpu_init_traps(struct pkvm_hyp_vcpu *hyp_vcpu)
 
 	pkvm_vcpu_reset_hcr(vcpu);
 
-	if ((!pkvm_hyp_vcpu_is_protected(hyp_vcpu)))
+	if ((!pkvm_hyp_vcpu_is_protected(hyp_vcpu))) {
+		struct kvm_vcpu *host_vcpu = hyp_vcpu->host_vcpu;
+
+		/* Trust the host for non-protected vcpu features. */
+		vcpu->arch.hcrx_el2 = host_vcpu->arch.hcrx_el2;
 		return 0;
+	}
 
 	ret = pkvm_check_pvm_cpu_features(vcpu);
 	if (ret)
@@ -176,6 +181,7 @@ static int pkvm_vcpu_init_traps(struct pkvm_hyp_vcpu *hyp_vcpu)
 
 	pvm_init_traps_hcr(vcpu);
 	pvm_init_traps_mdcr(vcpu);
+	vcpu_set_hcrx(vcpu);
 
 	return 0;
 }
@@ -240,10 +246,12 @@ struct pkvm_hyp_vcpu *pkvm_load_hyp_vcpu(pkvm_handle_t handle,
 
 	hyp_spin_lock(&vm_table_lock);
 	hyp_vm = get_vm_by_handle(handle);
-	if (!hyp_vm || hyp_vm->nr_vcpus <= vcpu_idx)
+	if (!hyp_vm || hyp_vm->kvm.created_vcpus <= vcpu_idx)
 		goto unlock;
 
 	hyp_vcpu = hyp_vm->vcpus[vcpu_idx];
+	if (!hyp_vcpu)
+		goto unlock;
 
 	/* Ensure vcpu isn't loaded on more than one cpu simultaneously. */
 	if (unlikely(hyp_vcpu->loaded_hyp_vcpu)) {
@@ -369,8 +377,14 @@ static void unpin_host_vcpus(struct pkvm_hyp_vcpu *hyp_vcpus[],
 {
 	int i;
 
-	for (i = 0; i < nr_vcpus; i++)
-		unpin_host_vcpu(hyp_vcpus[i]->host_vcpu);
+	for (i = 0; i < nr_vcpus; i++) {
+		struct pkvm_hyp_vcpu *hyp_vcpu = hyp_vcpus[i];
+
+		if (!hyp_vcpu)
+			continue;
+
+		unpin_host_vcpu(hyp_vcpu->host_vcpu);
+	}
 }
 
 static void init_pkvm_hyp_vm(struct kvm *host_kvm, struct pkvm_hyp_vm *hyp_vm,
@@ -394,24 +408,18 @@ static void pkvm_vcpu_init_sve(struct pkvm_hyp_vcpu *hyp_vcpu, struct kvm_vcpu *
 
 static int init_pkvm_hyp_vcpu(struct pkvm_hyp_vcpu *hyp_vcpu,
 			      struct pkvm_hyp_vm *hyp_vm,
-			      struct kvm_vcpu *host_vcpu,
-			      unsigned int vcpu_idx)
+			      struct kvm_vcpu *host_vcpu)
 {
 	int ret = 0;
 
 	if (hyp_pin_shared_mem(host_vcpu, host_vcpu + 1))
 		return -EBUSY;
 
-	if (host_vcpu->vcpu_idx != vcpu_idx) {
-		ret = -EINVAL;
-		goto done;
-	}
-
 	hyp_vcpu->host_vcpu = host_vcpu;
 
 	hyp_vcpu->vcpu.kvm = &hyp_vm->kvm;
 	hyp_vcpu->vcpu.vcpu_id = READ_ONCE(host_vcpu->vcpu_id);
-	hyp_vcpu->vcpu.vcpu_idx = vcpu_idx;
+	hyp_vcpu->vcpu.vcpu_idx = READ_ONCE(host_vcpu->vcpu_idx);
 
 	hyp_vcpu->vcpu.arch.hw_mmu = &hyp_vm->kvm.arch.mmu;
 	hyp_vcpu->vcpu.arch.cflags = READ_ONCE(host_vcpu->arch.cflags);
@@ -649,27 +657,28 @@ int __pkvm_init_vcpu(pkvm_handle_t handle, struct kvm_vcpu *host_vcpu,
 		goto unlock;
 	}
 
-	idx = hyp_vm->nr_vcpus;
+	ret = init_pkvm_hyp_vcpu(hyp_vcpu, hyp_vm, host_vcpu);
+	if (ret)
+		goto unlock;
+
+	idx = hyp_vcpu->vcpu.vcpu_idx;
 	if (idx >= hyp_vm->kvm.created_vcpus) {
 		ret = -EINVAL;
 		goto unlock;
 	}
 
-	ret = init_pkvm_hyp_vcpu(hyp_vcpu, hyp_vm, host_vcpu, idx);
-	if (ret)
+	if (hyp_vm->vcpus[idx]) {
+		ret = -EINVAL;
 		goto unlock;
+	}
 
 	hyp_vm->vcpus[idx] = hyp_vcpu;
-	hyp_vm->nr_vcpus++;
 unlock:
 	hyp_spin_unlock(&vm_table_lock);
 
-	if (ret) {
+	if (ret)
 		unmap_donated_memory(hyp_vcpu, sizeof(*hyp_vcpu));
-		return ret;
-	}
-
-	return 0;
+	return ret;
 }
 
 static void
@@ -686,7 +695,7 @@ teardown_donated_memory(struct kvm_hyp_memcache *mc, void *addr, size_t size)
 
 int __pkvm_teardown_vm(pkvm_handle_t handle)
 {
-	struct kvm_hyp_memcache *mc;
+	struct kvm_hyp_memcache *mc, *stage2_mc;
 	struct pkvm_hyp_vm *hyp_vm;
 	struct kvm *host_kvm;
 	unsigned int idx;
@@ -714,18 +723,24 @@ int __pkvm_teardown_vm(pkvm_handle_t handle)
 
 	/* Reclaim guest pages (including page-table pages) */
 	mc = &host_kvm->arch.pkvm.teardown_mc;
-	reclaim_guest_pages(hyp_vm, mc);
-	unpin_host_vcpus(hyp_vm->vcpus, hyp_vm->nr_vcpus);
+	stage2_mc = &host_kvm->arch.pkvm.stage2_teardown_mc;
+	reclaim_pgtable_pages(hyp_vm, stage2_mc);
+	unpin_host_vcpus(hyp_vm->vcpus, hyp_vm->kvm.created_vcpus);
 
 	/* Push the metadata pages to the teardown memcache */
-	for (idx = 0; idx < hyp_vm->nr_vcpus; ++idx) {
+	for (idx = 0; idx < hyp_vm->kvm.created_vcpus; ++idx) {
 		struct pkvm_hyp_vcpu *hyp_vcpu = hyp_vm->vcpus[idx];
-		struct kvm_hyp_memcache *vcpu_mc = &hyp_vcpu->vcpu.arch.pkvm_memcache;
+		struct kvm_hyp_memcache *vcpu_mc;
+
+		if (!hyp_vcpu)
+			continue;
+
+		vcpu_mc = &hyp_vcpu->vcpu.arch.pkvm_memcache;
 
 		while (vcpu_mc->nr_pages) {
 			void *addr = pop_hyp_memcache(vcpu_mc, hyp_phys_to_virt);
 
-			push_hyp_memcache(mc, addr, hyp_virt_to_phys);
+			push_hyp_memcache(stage2_mc, addr, hyp_virt_to_phys);
 			unmap_donated_memory_noclear(addr, PAGE_SIZE);
 		}
 
diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c
index 1f55b0c7b11d..2feb6c6b63af 100644
--- a/arch/arm64/kvm/mmu.c
+++ b/arch/arm64/kvm/mmu.c
@@ -1086,14 +1086,26 @@ void kvm_free_stage2_pgd(struct kvm_s2_mmu *mmu)
 	}
 }
 
-static void hyp_mc_free_fn(void *addr, void *unused)
+static void hyp_mc_free_fn(void *addr, void *mc)
 {
+	struct kvm_hyp_memcache *memcache = mc;
+
+	if (memcache->flags & HYP_MEMCACHE_ACCOUNT_STAGE2)
+		kvm_account_pgtable_pages(addr, -1);
+
 	free_page((unsigned long)addr);
 }
 
-static void *hyp_mc_alloc_fn(void *unused)
+static void *hyp_mc_alloc_fn(void *mc)
 {
-	return (void *)__get_free_page(GFP_KERNEL_ACCOUNT);
+	struct kvm_hyp_memcache *memcache = mc;
+	void *addr;
+
+	addr = (void *)__get_free_page(GFP_KERNEL_ACCOUNT);
+	if (addr && memcache->flags & HYP_MEMCACHE_ACCOUNT_STAGE2)
+		kvm_account_pgtable_pages(addr, 1);
+
+	return addr;
 }
 
 void free_hyp_memcache(struct kvm_hyp_memcache *mc)
@@ -1102,7 +1114,7 @@ void free_hyp_memcache(struct kvm_hyp_memcache *mc)
 		return;
 
 	kfree(mc->mapping);
-	__free_hyp_memcache(mc, hyp_mc_free_fn, kvm_host_va, NULL);
+	__free_hyp_memcache(mc, hyp_mc_free_fn, kvm_host_va, mc);
 }
 
 int topup_hyp_memcache(struct kvm_hyp_memcache *mc, unsigned long min_pages)
@@ -1117,7 +1129,7 @@ int topup_hyp_memcache(struct kvm_hyp_memcache *mc, unsigned long min_pages)
 	}
 
 	return __topup_hyp_memcache(mc, min_pages, hyp_mc_alloc_fn,
-				    kvm_host_pa, NULL);
+				    kvm_host_pa, mc);
 }
 
 /**
diff --git a/arch/arm64/kvm/pkvm.c b/arch/arm64/kvm/pkvm.c
index 930b677eb9b0..0f89157d31fd 100644
--- a/arch/arm64/kvm/pkvm.c
+++ b/arch/arm64/kvm/pkvm.c
@@ -111,6 +111,29 @@ static void __pkvm_destroy_hyp_vm(struct kvm *host_kvm)
 
 	host_kvm->arch.pkvm.handle = 0;
 	free_hyp_memcache(&host_kvm->arch.pkvm.teardown_mc);
+	free_hyp_memcache(&host_kvm->arch.pkvm.stage2_teardown_mc);
+}
+
+static int __pkvm_create_hyp_vcpu(struct kvm_vcpu *vcpu)
+{
+	size_t hyp_vcpu_sz = PAGE_ALIGN(PKVM_HYP_VCPU_SIZE);
+	pkvm_handle_t handle = vcpu->kvm->arch.pkvm.handle;
+	void *hyp_vcpu;
+	int ret;
+
+	vcpu->arch.pkvm_memcache.flags |= HYP_MEMCACHE_ACCOUNT_STAGE2;
+
+	hyp_vcpu = alloc_pages_exact(hyp_vcpu_sz, GFP_KERNEL_ACCOUNT);
+	if (!hyp_vcpu)
+		return -ENOMEM;
+
+	ret = kvm_call_hyp_nvhe(__pkvm_init_vcpu, handle, vcpu, hyp_vcpu);
+	if (!ret)
+		vcpu_set_flag(vcpu, VCPU_PKVM_FINALIZED);
+	else
+		free_pages_exact(hyp_vcpu, hyp_vcpu_sz);
+
+	return ret;
 }
 
 /*
@@ -125,11 +148,8 @@ static void __pkvm_destroy_hyp_vm(struct kvm *host_kvm)
  */
 static int __pkvm_create_hyp_vm(struct kvm *host_kvm)
 {
-	size_t pgd_sz, hyp_vm_sz, hyp_vcpu_sz;
-	struct kvm_vcpu *host_vcpu;
-	pkvm_handle_t handle;
+	size_t pgd_sz, hyp_vm_sz;
 	void *pgd, *hyp_vm;
-	unsigned long idx;
 	int ret;
 
 	if (host_kvm->created_vcpus < 1)
@@ -161,40 +181,11 @@ static int __pkvm_create_hyp_vm(struct kvm *host_kvm)
 	if (ret < 0)
 		goto free_vm;
 
-	handle = ret;
-
-	host_kvm->arch.pkvm.handle = handle;
-
-	/* Donate memory for the vcpus at hyp and initialize it. */
-	hyp_vcpu_sz = PAGE_ALIGN(PKVM_HYP_VCPU_SIZE);
-	kvm_for_each_vcpu(idx, host_vcpu, host_kvm) {
-		void *hyp_vcpu;
-
-		/* Indexing of the vcpus to be sequential starting at 0. */
-		if (WARN_ON(host_vcpu->vcpu_idx != idx)) {
-			ret = -EINVAL;
-			goto destroy_vm;
-		}
-
-		hyp_vcpu = alloc_pages_exact(hyp_vcpu_sz, GFP_KERNEL_ACCOUNT);
-		if (!hyp_vcpu) {
-			ret = -ENOMEM;
-			goto destroy_vm;
-		}
-
-		ret = kvm_call_hyp_nvhe(__pkvm_init_vcpu, handle, host_vcpu,
-					hyp_vcpu);
-		if (ret) {
-			free_pages_exact(hyp_vcpu, hyp_vcpu_sz);
-			goto destroy_vm;
-		}
-	}
+	host_kvm->arch.pkvm.handle = ret;
+	host_kvm->arch.pkvm.stage2_teardown_mc.flags |= HYP_MEMCACHE_ACCOUNT_STAGE2;
+	kvm_account_pgtable_pages(pgd, pgd_sz / PAGE_SIZE);
 
 	return 0;
-
-destroy_vm:
-	__pkvm_destroy_hyp_vm(host_kvm);
-	return ret;
 free_vm:
 	free_pages_exact(hyp_vm, hyp_vm_sz);
 free_pgd:
@@ -214,6 +205,18 @@ int pkvm_create_hyp_vm(struct kvm *host_kvm)
 	return ret;
 }
 
+int pkvm_create_hyp_vcpu(struct kvm_vcpu *vcpu)
+{
+	int ret = 0;
+
+	mutex_lock(&vcpu->kvm->arch.config_lock);
+	if (!vcpu_get_flag(vcpu, VCPU_PKVM_FINALIZED))
+		ret = __pkvm_create_hyp_vcpu(vcpu);
+	mutex_unlock(&vcpu->kvm->arch.config_lock);
+
+	return ret;
+}
+
 void pkvm_destroy_hyp_vm(struct kvm *host_kvm)
 {
 	mutex_lock(&host_kvm->arch.config_lock);
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 7435f9d152a9..ba0b63cd26dd 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -5132,25 +5132,7 @@ void kvm_calculate_traps(struct kvm_vcpu *vcpu)
 	mutex_lock(&kvm->arch.config_lock);
 	vcpu_set_hcr(vcpu);
 	vcpu_set_ich_hcr(vcpu);
-
-	if (cpus_have_final_cap(ARM64_HAS_HCX)) {
-		/*
-		 * In general, all HCRX_EL2 bits are gated by a feature.
-		 * The only reason we can set SMPME without checking any
-		 * feature is that its effects are not directly observable
-		 * from the guest.
-		 */
-		vcpu->arch.hcrx_el2 = HCRX_EL2_SMPME;
-
-		if (kvm_has_feat(kvm, ID_AA64ISAR2_EL1, MOPS, IMP))
-			vcpu->arch.hcrx_el2 |= (HCRX_EL2_MSCEn | HCRX_EL2_MCE2);
-
-		if (kvm_has_tcr2(kvm))
-			vcpu->arch.hcrx_el2 |= HCRX_EL2_TCR2En;
-
-		if (kvm_has_fpmr(kvm))
-			vcpu->arch.hcrx_el2 |= HCRX_EL2_EnFPM;
-	}
+	vcpu_set_hcrx(vcpu);
 
 	if (test_bit(KVM_ARCH_FLAG_FGU_INITIALIZED, &kvm->arch.flags))
 		goto out;