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

Pull kvm fixes from Paolo Bonzini:
"A bit late... I got sidetracked by back-from-vacation routines and
conferences. But most of these patches are already a few weeks old and
things look more calm on the mailing list than what this pull request
would suggest.

x86:

- missing TLB flush

- nested virtualization fixes for SMM (secure boot on nested
hypervisor) and other nested SVM fixes

- syscall fuzzing fixes

- live migration fix for AMD SEV

- mirror VMs now work for SEV-ES too

- fixes for reset

- possible out-of-bounds access in IOAPIC emulation

- fix enlightened VMCS on Windows 2022

ARM:

- Add missing FORCE target when building the EL2 object

- Fix a PMU probe regression on some platforms

Generic:

- KCSAN fixes

selftests:

- random fixes, mostly for clang compilation"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (43 commits)
selftests: KVM: Explicitly use movq to read xmm registers
selftests: KVM: Call ucall_init when setting up in rseq_test
KVM: Remove tlbs_dirty
KVM: X86: Synchronize the shadow pagetable before link it
KVM: X86: Fix missed remote tlb flush in rmap_write_protect()
KVM: x86: nSVM: don't copy virt_ext from vmcb12
KVM: x86: nSVM: test eax for 4K alignment for GP errata workaround
KVM: x86: selftests: test simultaneous uses of V_IRQ from L1 and L0
KVM: x86: nSVM: restore int_vector in svm_clear_vintr
kvm: x86: Add AMD PMU MSRs to msrs_to_save_all[]
KVM: x86: nVMX: re-evaluate emulation_required on nested VM exit
KVM: x86: nVMX: don't fail nested VM entry on invalid guest state if !from_vmentry
KVM: x86: VMX: synthesize invalid VM exit when emulating invalid guest state
KVM: x86: nSVM: refactor svm_leave_smm and smm_enter_smm
KVM: x86: SVM: call KVM_REQ_GET_NESTED_STATE_PAGES on exit from SMM mode
KVM: x86: reset pdptrs_from_userspace when exiting smm
KVM: x86: nSVM: restore the L1 host state prior to resuming nested guest on SMM exit
KVM: nVMX: Filter out all unsupported controls when eVMCS was activated
KVM: KVM: Use cpumask_available() to check for NULL cpumask when kicking vCPUs
KVM: Clean up benign vcpu->cpu data races when kicking vCPUs
...

arch/arm64/kvm/hyp/nvhe/Makefile

···
 #    runtime. Because the hypervisor is part of the kernel binary, relocations
 #    produce a kernel VA. We enumerate relocations targeting hyp at build time
 #    and convert the kernel VAs at those positions to hyp VAs.
-$(obj)/hyp-reloc.S: $(obj)/kvm_nvhe.tmp.o $(obj)/gen-hyprel
+$(obj)/hyp-reloc.S: $(obj)/kvm_nvhe.tmp.o $(obj)/gen-hyprel FORCE
 	$(call if_changed,hyprel)
 
 # 5) Compile hyp-reloc.S and link it into the existing partially linked object.

arch/arm64/kvm/perf.c

···
 
 int kvm_perf_init(void)
 {
-	if (kvm_pmu_probe_pmuver() != ID_AA64DFR0_PMUVER_IMP_DEF && !is_protected_kvm_enabled())
-		static_branch_enable(&kvm_arm_pmu_available);
-
 	return perf_register_guest_info_callbacks(&kvm_guest_cbs);
 }
 

arch/arm64/kvm/pmu-emul.c

···
 	kvm_pmu_create_perf_event(vcpu, select_idx);
 }
 
-int kvm_pmu_probe_pmuver(void)
+void kvm_host_pmu_init(struct arm_pmu *pmu)
+{
+	if (pmu->pmuver != 0 && pmu->pmuver != ID_AA64DFR0_PMUVER_IMP_DEF &&
+	    !kvm_arm_support_pmu_v3() && !is_protected_kvm_enabled())
+		static_branch_enable(&kvm_arm_pmu_available);
+}
+
+static int kvm_pmu_probe_pmuver(void)
 {
 	struct perf_event_attr attr = { };
 	struct perf_event *event;

arch/s390/kvm/interrupt.c

···
 static void __set_cpu_idle(struct kvm_vcpu *vcpu)
 {
 	kvm_s390_set_cpuflags(vcpu, CPUSTAT_WAIT);
-	set_bit(kvm_vcpu_get_idx(vcpu), vcpu->kvm->arch.idle_mask);
+	set_bit(vcpu->vcpu_idx, vcpu->kvm->arch.idle_mask);
 }
 
 static void __unset_cpu_idle(struct kvm_vcpu *vcpu)
 {
 	kvm_s390_clear_cpuflags(vcpu, CPUSTAT_WAIT);
-	clear_bit(kvm_vcpu_get_idx(vcpu), vcpu->kvm->arch.idle_mask);
+	clear_bit(vcpu->vcpu_idx, vcpu->kvm->arch.idle_mask);
 }
 
 static void __reset_intercept_indicators(struct kvm_vcpu *vcpu)

arch/s390/kvm/kvm-s390.c

···
 		kvm_s390_patch_guest_per_regs(vcpu);
 	}
 
-	clear_bit(kvm_vcpu_get_idx(vcpu), vcpu->kvm->arch.gisa_int.kicked_mask);
+	clear_bit(vcpu->vcpu_idx, vcpu->kvm->arch.gisa_int.kicked_mask);
 
 	vcpu->arch.sie_block->icptcode = 0;
 	cpuflags = atomic_read(&vcpu->arch.sie_block->cpuflags);

arch/s390/kvm/kvm-s390.h

···
 
 static inline int is_vcpu_idle(struct kvm_vcpu *vcpu)
 {
-	return test_bit(kvm_vcpu_get_idx(vcpu), vcpu->kvm->arch.idle_mask);
+	return test_bit(vcpu->vcpu_idx, vcpu->kvm->arch.idle_mask);
 }
 
 static inline int kvm_is_ucontrol(struct kvm *kvm)

arch/x86/include/asm/kvm_page_track.h

···
 			    struct kvm_page_track_notifier_node *node);
 };
 
-void kvm_page_track_init(struct kvm *kvm);
+int kvm_page_track_init(struct kvm *kvm);
 void kvm_page_track_cleanup(struct kvm *kvm);
 
 void kvm_page_track_free_memslot(struct kvm_memory_slot *slot);

arch/x86/kvm/emulate.c

···
 	u64 cr4 = ctxt->ops->get_cr(ctxt, 4);
 
 	if (cr4 & X86_CR4_TSD && ctxt->ops->cpl(ctxt))
-		return emulate_ud(ctxt);
+		return emulate_gp(ctxt, 0);
 
 	return X86EMUL_CONTINUE;
 }

arch/x86/kvm/hyperv.c

···
 	for (i = 0; i < ARRAY_SIZE(hv_vcpu->stimer); i++)
 		stimer_init(&hv_vcpu->stimer[i], i);
 
-	hv_vcpu->vp_index = kvm_vcpu_get_idx(vcpu);
+	hv_vcpu->vp_index = vcpu->vcpu_idx;
 
 	return 0;
 }
···
 	switch (msr) {
 	case HV_X64_MSR_VP_INDEX: {
 		struct kvm_hv *hv = to_kvm_hv(vcpu->kvm);
-		int vcpu_idx = kvm_vcpu_get_idx(vcpu);
 		u32 new_vp_index = (u32)data;
 
 		if (!host || new_vp_index >= KVM_MAX_VCPUS)
···
 		 * VP index is changing, adjust num_mismatched_vp_indexes if
 		 * it now matches or no longer matches vcpu_idx.
 		 */
-		if (hv_vcpu->vp_index == vcpu_idx)
+		if (hv_vcpu->vp_index == vcpu->vcpu_idx)
 			atomic_inc(&hv->num_mismatched_vp_indexes);
-		else if (new_vp_index == vcpu_idx)
+		else if (new_vp_index == vcpu->vcpu_idx)
 			atomic_dec(&hv->num_mismatched_vp_indexes);
 
 		hv_vcpu->vp_index = new_vp_index;

arch/x86/kvm/hyperv.h

···
 {
 	struct kvm_vcpu_hv *hv_vcpu = to_hv_vcpu(vcpu);
 
-	return hv_vcpu ? hv_vcpu->vp_index : kvm_vcpu_get_idx(vcpu);
+	return hv_vcpu ? hv_vcpu->vp_index : vcpu->vcpu_idx;
 }
 
 int kvm_hv_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data, bool host);

arch/x86/kvm/ioapic.c

···
 	unsigned index;
 	bool mask_before, mask_after;
 	union kvm_ioapic_redirect_entry *e;
-	unsigned long vcpu_bitmap;
 	int old_remote_irr, old_delivery_status, old_dest_id, old_dest_mode;
+	DECLARE_BITMAP(vcpu_bitmap, KVM_MAX_VCPUS);
 
 	switch (ioapic->ioregsel) {
 	case IOAPIC_REG_VERSION:
···
 			irq.shorthand = APIC_DEST_NOSHORT;
 			irq.dest_id = e->fields.dest_id;
 			irq.msi_redir_hint = false;
-			bitmap_zero(&vcpu_bitmap, 16);
+			bitmap_zero(vcpu_bitmap, KVM_MAX_VCPUS);
 			kvm_bitmap_or_dest_vcpus(ioapic->kvm, &irq,
-						 &vcpu_bitmap);
+						 vcpu_bitmap);
 			if (old_dest_mode != e->fields.dest_mode ||
 			    old_dest_id != e->fields.dest_id) {
 				/*
···
 				    kvm_lapic_irq_dest_mode(
 					!!e->fields.dest_mode);
 				kvm_bitmap_or_dest_vcpus(ioapic->kvm, &irq,
-							 &vcpu_bitmap);
+							 vcpu_bitmap);
 			}
 			kvm_make_scan_ioapic_request_mask(ioapic->kvm,
-							  &vcpu_bitmap);
+							  vcpu_bitmap);
 		} else {
 			kvm_make_scan_ioapic_request(ioapic->kvm);
 		}

arch/x86/kvm/mmu/mmu.c

···
 	} while (!sp->unsync_children);
 }
 
-static void mmu_sync_children(struct kvm_vcpu *vcpu,
-			      struct kvm_mmu_page *parent)
+static int mmu_sync_children(struct kvm_vcpu *vcpu,
+			     struct kvm_mmu_page *parent, bool can_yield)
 {
 	int i;
 	struct kvm_mmu_page *sp;
···
 		}
 		if (need_resched() || rwlock_needbreak(&vcpu->kvm->mmu_lock)) {
 			kvm_mmu_flush_or_zap(vcpu, &invalid_list, false, flush);
+			if (!can_yield) {
+				kvm_make_request(KVM_REQ_MMU_SYNC, vcpu);
+				return -EINTR;
+			}
+
 			cond_resched_rwlock_write(&vcpu->kvm->mmu_lock);
 			flush = false;
 		}
 	}
 
 	kvm_mmu_flush_or_zap(vcpu, &invalid_list, false, flush);
+	return 0;
 }
 
 static void __clear_sp_write_flooding_count(struct kvm_mmu_page *sp)
···
 			WARN_ON(!list_empty(&invalid_list));
 			kvm_make_request(KVM_REQ_TLB_FLUSH_CURRENT, vcpu);
 		}
-
-		if (sp->unsync_children)
-			kvm_make_request(KVM_REQ_MMU_SYNC, vcpu);
 
 		__clear_sp_write_flooding_count(sp);
 
···
 		write_lock(&vcpu->kvm->mmu_lock);
 		kvm_mmu_audit(vcpu, AUDIT_PRE_SYNC);
 
-		mmu_sync_children(vcpu, sp);
+		mmu_sync_children(vcpu, sp, true);
 
 		kvm_mmu_audit(vcpu, AUDIT_POST_SYNC);
 		write_unlock(&vcpu->kvm->mmu_lock);
···
 		if (IS_VALID_PAE_ROOT(root)) {
 			root &= PT64_BASE_ADDR_MASK;
 			sp = to_shadow_page(root);
-			mmu_sync_children(vcpu, sp);
+			mmu_sync_children(vcpu, sp, true);
 		}
 	}
 

arch/x86/kvm/mmu/page_track.c

···
 	cleanup_srcu_struct(&head->track_srcu);
 }
 
-void kvm_page_track_init(struct kvm *kvm)
+int kvm_page_track_init(struct kvm *kvm)
 {
 	struct kvm_page_track_notifier_head *head;
 
 	head = &kvm->arch.track_notifier_head;
-	init_srcu_struct(&head->track_srcu);
 	INIT_HLIST_HEAD(&head->track_notifier_list);
+	return init_srcu_struct(&head->track_srcu);
 }
 
 /*

arch/x86/kvm/mmu/paging_tmpl.h

···
 		if (!is_shadow_present_pte(*it.sptep)) {
 			table_gfn = gw->table_gfn[it.level - 2];
 			access = gw->pt_access[it.level - 2];
-			sp = kvm_mmu_get_page(vcpu, table_gfn, addr, it.level-1,
-					      false, access);
+			sp = kvm_mmu_get_page(vcpu, table_gfn, addr,
+					      it.level-1, false, access);
+			/*
+			 * We must synchronize the pagetable before linking it
+			 * because the guest doesn't need to flush tlb when
+			 * the gpte is changed from non-present to present.
+			 * Otherwise, the guest may use the wrong mapping.
+			 *
+			 * For PG_LEVEL_4K, kvm_mmu_get_page() has already
+			 * synchronized it transiently via kvm_sync_page().
+			 *
+			 * For higher level pagetable, we synchronize it via
+			 * the slower mmu_sync_children().  If it needs to
+			 * break, some progress has been made; return
+			 * RET_PF_RETRY and retry on the next #PF.
+			 * KVM_REQ_MMU_SYNC is not necessary but it
+			 * expedites the process.
+			 */
+			if (sp->unsync_children &&
+			    mmu_sync_children(vcpu, sp, false))
+				return RET_PF_RETRY;
 		}
 
 		/*
···
  * Using the cached information from sp->gfns is safe because:
  * - The spte has a reference to the struct page, so the pfn for a given gfn
  *   can't change unless all sptes pointing to it are nuked first.
- *
- * Note:
- *   We should flush all tlbs if spte is dropped even though guest is
- *   responsible for it. Since if we don't, kvm_mmu_notifier_invalidate_page
- *   and kvm_mmu_notifier_invalidate_range_start detect the mapping page isn't
- *   used by guest then tlbs are not flushed, so guest is allowed to access the
- *   freed pages.
- *   And we increase kvm->tlbs_dirty to delay tlbs flush in this case.
  */
 static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 {
···
 			return 0;
 
 		if (FNAME(prefetch_invalid_gpte)(vcpu, sp, &sp->spt[i], gpte)) {
-			/*
-			 * Update spte before increasing tlbs_dirty to make
-			 * sure no tlb flush is lost after spte is zapped; see
-			 * the comments in kvm_flush_remote_tlbs().
-			 */
-			smp_wmb();
-			vcpu->kvm->tlbs_dirty++;
+			set_spte_ret |= SET_SPTE_NEED_REMOTE_TLB_FLUSH;
 			continue;
 		}
 
···
 
 		if (gfn != sp->gfns[i]) {
 			drop_spte(vcpu->kvm, &sp->spt[i]);
-			/*
-			 * The same as above where we are doing
-			 * prefetch_invalid_gpte().
-			 */
-			smp_wmb();
-			vcpu->kvm->tlbs_dirty++;
+			set_spte_ret |= SET_SPTE_NEED_REMOTE_TLB_FLUSH;
 			continue;
 		}
 

arch/x86/kvm/svm/nested.c

···
 		(svm->nested.ctl.int_ctl & int_ctl_vmcb12_bits) |
 		(svm->vmcb01.ptr->control.int_ctl & int_ctl_vmcb01_bits);
 
-	svm->vmcb->control.virt_ext            = svm->nested.ctl.virt_ext;
 	svm->vmcb->control.int_vector          = svm->nested.ctl.int_vector;
 	svm->vmcb->control.int_state           = svm->nested.ctl.int_state;
 	svm->vmcb->control.event_inj           = svm->nested.ctl.event_inj;
···
 }
 
 int enter_svm_guest_mode(struct kvm_vcpu *vcpu, u64 vmcb12_gpa,
-			 struct vmcb *vmcb12)
+			 struct vmcb *vmcb12, bool from_vmrun)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 	int ret;
···
 	nested_vmcb02_prepare_save(svm, vmcb12);
 
 	ret = nested_svm_load_cr3(&svm->vcpu, vmcb12->save.cr3,
-				  nested_npt_enabled(svm), true);
+				  nested_npt_enabled(svm), from_vmrun);
 	if (ret)
 		return ret;
 
 	if (!npt_enabled)
 		vcpu->arch.mmu->inject_page_fault = svm_inject_page_fault_nested;
+
+	if (!from_vmrun)
+		kvm_make_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu);
 
 	svm_set_gif(svm, true);
 
···
 
 	svm->nested.nested_run_pending = 1;
 
-	if (enter_svm_guest_mode(vcpu, vmcb12_gpa, vmcb12))
+	if (enter_svm_guest_mode(vcpu, vmcb12_gpa, vmcb12, true))
 		goto out_exit_err;
 
 	if (nested_svm_vmrun_msrpm(svm))

arch/x86/kvm/svm/sev.c

···
 	return 0;
 }
 
+static int __sev_launch_update_vmsa(struct kvm *kvm, struct kvm_vcpu *vcpu,
+				    int *error)
+{
+	struct sev_data_launch_update_vmsa vmsa;
+	struct vcpu_svm *svm = to_svm(vcpu);
+	int ret;
+
+	/* Perform some pre-encryption checks against the VMSA */
+	ret = sev_es_sync_vmsa(svm);
+	if (ret)
+		return ret;
+
+	/*
+	 * The LAUNCH_UPDATE_VMSA command will perform in-place encryption of
+	 * the VMSA memory content (i.e it will write the same memory region
+	 * with the guest's key), so invalidate it first.
+	 */
+	clflush_cache_range(svm->vmsa, PAGE_SIZE);
+
+	vmsa.reserved = 0;
+	vmsa.handle = to_kvm_svm(kvm)->sev_info.handle;
+	vmsa.address = __sme_pa(svm->vmsa);
+	vmsa.len = PAGE_SIZE;
+	return sev_issue_cmd(kvm, SEV_CMD_LAUNCH_UPDATE_VMSA, &vmsa, error);
+}
+
 static int sev_launch_update_vmsa(struct kvm *kvm, struct kvm_sev_cmd *argp)
 {
-	struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
-	struct sev_data_launch_update_vmsa vmsa;
 	struct kvm_vcpu *vcpu;
 	int i, ret;
 
 	if (!sev_es_guest(kvm))
 		return -ENOTTY;
 
-	vmsa.reserved = 0;
-
 	kvm_for_each_vcpu(i, vcpu, kvm) {
-		struct vcpu_svm *svm = to_svm(vcpu);
-
-		/* Perform some pre-encryption checks against the VMSA */
-		ret = sev_es_sync_vmsa(svm);
+		ret = mutex_lock_killable(&vcpu->mutex);
 		if (ret)
 			return ret;
 
-		/*
-		 * The LAUNCH_UPDATE_VMSA command will perform in-place
-		 * encryption of the VMSA memory content (i.e it will write
-		 * the same memory region with the guest's key), so invalidate
-		 * it first.
-		 */
-		clflush_cache_range(svm->vmsa, PAGE_SIZE);
+		ret = __sev_launch_update_vmsa(kvm, vcpu, &argp->error);
 
-		vmsa.handle = sev->handle;
-		vmsa.address = __sme_pa(svm->vmsa);
-		vmsa.len = PAGE_SIZE;
-		ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_UPDATE_VMSA, &vmsa,
-				    &argp->error);
+		mutex_unlock(&vcpu->mutex);
 		if (ret)
 			return ret;
-
-		svm->vcpu.arch.guest_state_protected = true;
 	}
 
 	return 0;
···
 
 	/* Bind ASID to this guest */
 	ret = sev_bind_asid(kvm, start.handle, error);
-	if (ret)
+	if (ret) {
+		sev_decommission(start.handle);
 		goto e_free_session;
+	}
 
 	params.handle = start.handle;
 	if (copy_to_user((void __user *)(uintptr_t)argp->data,
···
 
 	/* Pin guest memory */
 	guest_page = sev_pin_memory(kvm, params.guest_uaddr & PAGE_MASK,
-				    PAGE_SIZE, &n, 0);
+				    PAGE_SIZE, &n, 1);
 	if (IS_ERR(guest_page)) {
 		ret = PTR_ERR(guest_page);
 		goto e_free_trans;
···
 	return sev_issue_cmd(kvm, SEV_CMD_RECEIVE_FINISH, &data, &argp->error);
 }
 
+static bool cmd_allowed_from_miror(u32 cmd_id)
+{
+	/*
+	 * Allow mirrors VM to call KVM_SEV_LAUNCH_UPDATE_VMSA to enable SEV-ES
+	 * active mirror VMs. Also allow the debugging and status commands.
+	 */
+	if (cmd_id == KVM_SEV_LAUNCH_UPDATE_VMSA ||
+	    cmd_id == KVM_SEV_GUEST_STATUS || cmd_id == KVM_SEV_DBG_DECRYPT ||
+	    cmd_id == KVM_SEV_DBG_ENCRYPT)
+		return true;
+
+	return false;
+}
+
 int svm_mem_enc_op(struct kvm *kvm, void __user *argp)
 {
 	struct kvm_sev_cmd sev_cmd;
···
 
 	mutex_lock(&kvm->lock);
 
-	/* enc_context_owner handles all memory enc operations */
-	if (is_mirroring_enc_context(kvm)) {
+	/* Only the enc_context_owner handles some memory enc operations. */
+	if (is_mirroring_enc_context(kvm) &&
+	    !cmd_allowed_from_miror(sev_cmd.id)) {
 		r = -EINVAL;
 		goto out;
 	}
···
 {
 	struct file *source_kvm_file;
 	struct kvm *source_kvm;
-	struct kvm_sev_info *mirror_sev;
-	unsigned int asid;
+	struct kvm_sev_info source_sev, *mirror_sev;
 	int ret;
 
 	source_kvm_file = fget(source_fd);
···
 		goto e_source_unlock;
 	}
 
-	asid = to_kvm_svm(source_kvm)->sev_info.asid;
+	memcpy(&source_sev, &to_kvm_svm(source_kvm)->sev_info,
+	       sizeof(source_sev));
 
 	/*
 	 * The mirror kvm holds an enc_context_owner ref so its asid can't
···
 	/* Set enc_context_owner and copy its encryption context over */
 	mirror_sev = &to_kvm_svm(kvm)->sev_info;
 	mirror_sev->enc_context_owner = source_kvm;
-	mirror_sev->asid = asid;
 	mirror_sev->active = true;
+	mirror_sev->asid = source_sev.asid;
+	mirror_sev->fd = source_sev.fd;
+	mirror_sev->es_active = source_sev.es_active;
+	mirror_sev->handle = source_sev.handle;
+	/*
+	 * Do not copy ap_jump_table. Since the mirror does not share the same
+	 * KVM contexts as the original, and they may have different
+	 * memory-views.
+	 */
 
 	mutex_unlock(&kvm->lock);
 	return 0;

arch/x86/kvm/svm/svm.c

···
 
 		svm->vmcb->control.int_ctl |= svm->nested.ctl.int_ctl &
 			V_IRQ_INJECTION_BITS_MASK;
+
+		svm->vmcb->control.int_vector = svm->nested.ctl.int_vector;
 	}
 
 	vmcb_mark_dirty(svm->vmcb, VMCB_INTR);
···
 
 	/* Both #GP cases have zero error_code */
 	if (error_code)
+		goto reinject;
+
+	/* All SVM instructions expect page aligned RAX */
+	if (svm->vmcb->save.rax & ~PAGE_MASK)
 		goto reinject;
 
 	/* Decode the instruction for usage later */
···
 	struct kvm_host_map map_save;
 	int ret;
 
-	if (is_guest_mode(vcpu)) {
-		/* FED8h - SVM Guest */
-		put_smstate(u64, smstate, 0x7ed8, 1);
-		/* FEE0h - SVM Guest VMCB Physical Address */
-		put_smstate(u64, smstate, 0x7ee0, svm->nested.vmcb12_gpa);
+	if (!is_guest_mode(vcpu))
+		return 0;
 
-		svm->vmcb->save.rax = vcpu->arch.regs[VCPU_REGS_RAX];
-		svm->vmcb->save.rsp = vcpu->arch.regs[VCPU_REGS_RSP];
-		svm->vmcb->save.rip = vcpu->arch.regs[VCPU_REGS_RIP];
+	/* FED8h - SVM Guest */
+	put_smstate(u64, smstate, 0x7ed8, 1);
+	/* FEE0h - SVM Guest VMCB Physical Address */
+	put_smstate(u64, smstate, 0x7ee0, svm->nested.vmcb12_gpa);
 
-		ret = nested_svm_vmexit(svm);
-		if (ret)
-			return ret;
+	svm->vmcb->save.rax = vcpu->arch.regs[VCPU_REGS_RAX];
+	svm->vmcb->save.rsp = vcpu->arch.regs[VCPU_REGS_RSP];
+	svm->vmcb->save.rip = vcpu->arch.regs[VCPU_REGS_RIP];
 
-		/*
-		 * KVM uses VMCB01 to store L1 host state while L2 runs but
-		 * VMCB01 is going to be used during SMM and thus the state will
-		 * be lost. Temporary save non-VMLOAD/VMSAVE state to the host save
-		 * area pointed to by MSR_VM_HSAVE_PA. APM guarantees that the
-		 * format of the area is identical to guest save area offsetted
-		 * by 0x400 (matches the offset of 'struct vmcb_save_area'
-		 * within 'struct vmcb'). Note: HSAVE area may also be used by
-		 * L1 hypervisor to save additional host context (e.g. KVM does
-		 * that, see svm_prepare_guest_switch()) which must be
-		 * preserved.
-		 */
-		if (kvm_vcpu_map(vcpu, gpa_to_gfn(svm->nested.hsave_msr),
-				 &map_save) == -EINVAL)
-			return 1;
+	ret = nested_svm_vmexit(svm);
+	if (ret)
+		return ret;
 
-		BUILD_BUG_ON(offsetof(struct vmcb, save) != 0x400);
+	/*
+	 * KVM uses VMCB01 to store L1 host state while L2 runs but
+	 * VMCB01 is going to be used during SMM and thus the state will
+	 * be lost. Temporary save non-VMLOAD/VMSAVE state to the host save
+	 * area pointed to by MSR_VM_HSAVE_PA. APM guarantees that the
+	 * format of the area is identical to guest save area offsetted
+	 * by 0x400 (matches the offset of 'struct vmcb_save_area'
+	 * within 'struct vmcb'). Note: HSAVE area may also be used by
+	 * L1 hypervisor to save additional host context (e.g. KVM does
+	 * that, see svm_prepare_guest_switch()) which must be
+	 * preserved.
+	 */
+	if (kvm_vcpu_map(vcpu, gpa_to_gfn(svm->nested.hsave_msr),
+			 &map_save) == -EINVAL)
+		return 1;
 
-		svm_copy_vmrun_state(map_save.hva + 0x400,
-				     &svm->vmcb01.ptr->save);
+	BUILD_BUG_ON(offsetof(struct vmcb, save) != 0x400);
 
-		kvm_vcpu_unmap(vcpu, &map_save, true);
-	}
+	svm_copy_vmrun_state(map_save.hva + 0x400,
+			     &svm->vmcb01.ptr->save);
+
+	kvm_vcpu_unmap(vcpu, &map_save, true);
 	return 0;
 }
 
···
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 	struct kvm_host_map map, map_save;
-	int ret = 0;
+	u64 saved_efer, vmcb12_gpa;
+	struct vmcb *vmcb12;
+	int ret;
 
-	if (guest_cpuid_has(vcpu, X86_FEATURE_LM)) {
-		u64 saved_efer = GET_SMSTATE(u64, smstate, 0x7ed0);
-		u64 guest = GET_SMSTATE(u64, smstate, 0x7ed8);
-		u64 vmcb12_gpa = GET_SMSTATE(u64, smstate, 0x7ee0);
-		struct vmcb *vmcb12;
+	if (!guest_cpuid_has(vcpu, X86_FEATURE_LM))
+		return 0;
 
-		if (guest) {
-			if (!guest_cpuid_has(vcpu, X86_FEATURE_SVM))
-				return 1;
+	/* Non-zero if SMI arrived while vCPU was in guest mode. */
+	if (!GET_SMSTATE(u64, smstate, 0x7ed8))
+		return 0;
 
-			if (!(saved_efer & EFER_SVME))
-				return 1;
+	if (!guest_cpuid_has(vcpu, X86_FEATURE_SVM))
+		return 1;
 
-			if (kvm_vcpu_map(vcpu,
-					 gpa_to_gfn(vmcb12_gpa), &map) == -EINVAL)
-				return 1;
+	saved_efer = GET_SMSTATE(u64, smstate, 0x7ed0);
+	if (!(saved_efer & EFER_SVME))
+		return 1;
 
-			if (svm_allocate_nested(svm))
-				return 1;
+	vmcb12_gpa = GET_SMSTATE(u64, smstate, 0x7ee0);
+	if (kvm_vcpu_map(vcpu, gpa_to_gfn(vmcb12_gpa), &map) == -EINVAL)
+		return 1;
 
-			vmcb12 = map.hva;
+	ret = 1;
+	if (kvm_vcpu_map(vcpu, gpa_to_gfn(svm->nested.hsave_msr), &map_save) == -EINVAL)
+		goto unmap_map;
 
-			nested_load_control_from_vmcb12(svm, &vmcb12->control);
+	if (svm_allocate_nested(svm))
+		goto unmap_save;
 
-			ret = enter_svm_guest_mode(vcpu, vmcb12_gpa, vmcb12);
-			kvm_vcpu_unmap(vcpu, &map, true);
+	/*
+	 * Restore L1 host state from L1 HSAVE area as VMCB01 was
+	 * used during SMM (see svm_enter_smm())
+	 */
 
-			/*
-			 * Restore L1 host state from L1 HSAVE area as VMCB01 was
-			 * used during SMM (see svm_enter_smm())
-			 */
-			if (kvm_vcpu_map(vcpu, gpa_to_gfn(svm->nested.hsave_msr),
-					 &map_save) == -EINVAL)
-				return 1;
+	svm_copy_vmrun_state(&svm->vmcb01.ptr->save, map_save.hva + 0x400);
 
-			svm_copy_vmrun_state(&svm->vmcb01.ptr->save,
-					     map_save.hva + 0x400);
+	/*
+	 * Enter the nested guest now
+	 */
 
-			kvm_vcpu_unmap(vcpu, &map_save, true);
-		}
-	}
+	vmcb12 = map.hva;
+	nested_load_control_from_vmcb12(svm, &vmcb12->control);
+	ret = enter_svm_guest_mode(vcpu, vmcb12_gpa, vmcb12, false);
 
+unmap_save:
+	kvm_vcpu_unmap(vcpu, &map_save, true);
+unmap_map:
+	kvm_vcpu_unmap(vcpu, &map, true);
 	return ret;
 }
 

arch/x86/kvm/svm/svm.h

···
 	return vmcb_is_intercept(&svm->nested.ctl, INTERCEPT_NMI);
 }
 
-int enter_svm_guest_mode(struct kvm_vcpu *vcpu, u64 vmcb_gpa, struct vmcb *vmcb12);
+int enter_svm_guest_mode(struct kvm_vcpu *vcpu,
+			 u64 vmcb_gpa, struct vmcb *vmcb12, bool from_vmrun);
 void svm_leave_nested(struct vcpu_svm *svm);
 void svm_free_nested(struct vcpu_svm *svm);
 int svm_allocate_nested(struct vcpu_svm *svm);

arch/x86/kvm/vmx/evmcs.c

···
 	switch (msr_index) {
 	case MSR_IA32_VMX_EXIT_CTLS:
 	case MSR_IA32_VMX_TRUE_EXIT_CTLS:
-		ctl_high &= ~VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL;
+		ctl_high &= ~EVMCS1_UNSUPPORTED_VMEXIT_CTRL;
 		break;
 	case MSR_IA32_VMX_ENTRY_CTLS:
 	case MSR_IA32_VMX_TRUE_ENTRY_CTLS:
-		ctl_high &= ~VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL;
+		ctl_high &= ~EVMCS1_UNSUPPORTED_VMENTRY_CTRL;
 		break;
 	case MSR_IA32_VMX_PROCBASED_CTLS2:
-		ctl_high &= ~SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES;
+		ctl_high &= ~EVMCS1_UNSUPPORTED_2NDEXEC;
+		break;
+	case MSR_IA32_VMX_PINBASED_CTLS:
+		ctl_high &= ~EVMCS1_UNSUPPORTED_PINCTRL;
+		break;
+	case MSR_IA32_VMX_VMFUNC:
+		ctl_low &= ~EVMCS1_UNSUPPORTED_VMFUNC;
 		break;
 	}
 

arch/x86/kvm/vmx/nested.c

···
 	 * Guest state is invalid and unrestricted guest is disabled,
 	 * which means L1 attempted VMEntry to L2 with invalid state.
 	 * Fail the VMEntry.
+	 *
+	 * However when force loading the guest state (SMM exit or
+	 * loading nested state after migration, it is possible to
+	 * have invalid guest state now, which will be later fixed by
+	 * restoring L2 register state
 	 */
-	if (CC(!vmx_guest_state_valid(vcpu))) {
+	if (CC(from_vmentry && !vmx_guest_state_valid(vcpu))) {
 		*entry_failure_code = ENTRY_FAIL_DEFAULT;
 		return -EINVAL;
 	}
···
 	if (nested_vmx_load_msr(vcpu, vmcs12->vm_exit_msr_load_addr,
 				vmcs12->vm_exit_msr_load_count))
 		nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_MSR_FAIL);
+
+	to_vmx(vcpu)->emulation_required = vmx_emulation_required(vcpu);
 }
 
 static inline u64 nested_vmx_get_vmcs01_guest_efer(struct vcpu_vmx *vmx)
···
 	return -ENOMEM;
 }
 
-/*
- * Emulate the VMXON instruction.
- * Currently, we just remember that VMX is active, and do not save or even
- * inspect the argument to VMXON (the so-called "VMXON pointer") because we
- * do not currently need to store anything in that guest-allocated memory
- * region. Consequently, VMCLEAR and VMPTRLD also do not verify that the their
- * argument is different from the VMXON pointer (which the spec says they do).
- */
+/* Emulate the VMXON instruction. */
 static int handle_vmon(struct kvm_vcpu *vcpu)
 {
 	int ret;
···
 		return true;
 	case EXIT_REASON_VMFUNC:
 		/* VM functions are emulated through L2->L0 vmexits. */
+		return true;
+	case EXIT_REASON_BUS_LOCK:
+		/*
+		 * At present, bus lock VM exit is never exposed to L1.
+		 * Handle L2's bus locks in L0 directly.
+		 */
 		return true;
 	default:
 		break;

arch/x86/kvm/vmx/vmx.c

···
 	vmx_prepare_switch_to_host(to_vmx(vcpu));
 }
 
-static bool emulation_required(struct kvm_vcpu *vcpu)
+bool vmx_emulation_required(struct kvm_vcpu *vcpu)
 {
 	return emulate_invalid_guest_state && !vmx_guest_state_valid(vcpu);
 }
···
 	vmcs_writel(GUEST_RFLAGS, rflags);
 
 	if ((old_rflags ^ vmx->rflags) & X86_EFLAGS_VM)
-		vmx->emulation_required = emulation_required(vcpu);
+		vmx->emulation_required = vmx_emulation_required(vcpu);
 }
 
 u32 vmx_get_interrupt_shadow(struct kvm_vcpu *vcpu)
···
 				    &msr_info->data))
 			return 1;
 		/*
-		 * Enlightened VMCS v1 doesn't have certain fields, but buggy
-		 * Hyper-V versions are still trying to use corresponding
-		 * features when they are exposed. Filter out the essential
-		 * minimum.
+		 * Enlightened VMCS v1 doesn't have certain VMCS fields but
+		 * instead of just ignoring the features, different Hyper-V
+		 * versions are either trying to use them and fail or do some
+		 * sanity checking and refuse to boot. Filter all unsupported
+		 * features out.
 		 */
 		if (!msr_info->host_initiated &&
 		    vmx->nested.enlightened_vmcs_enabled)
···
 	}
 
 	/* depends on vcpu->arch.cr0 to be set to a new value */
-	vmx->emulation_required = emulation_required(vcpu);
+	vmx->emulation_required = vmx_emulation_required(vcpu);
 }
 
 static int vmx_get_max_tdp_level(void)
···
 {
 	__vmx_set_segment(vcpu, var, seg);
 
-	to_vmx(vcpu)->emulation_required = emulation_required(vcpu);
+	to_vmx(vcpu)->emulation_required = vmx_emulation_required(vcpu);
 }
 
 static void vmx_get_cs_db_l_bits(struct kvm_vcpu *vcpu, int *db, int *l)
···
 		     vmx->loaded_vmcs->soft_vnmi_blocked))
 		vmx->loaded_vmcs->entry_time = ktime_get();
 
-	/* Don't enter VMX if guest state is invalid, let the exit handler
-	   start emulation until we arrive back to a valid state */
-	if (vmx->emulation_required)
+	/*
+	 * Don't enter VMX if guest state is invalid, let the exit handler
+	 * start emulation until we arrive back to a valid state.  Synthesize a
+	 * consistency check VM-Exit due to invalid guest state and bail.
+	 */
+	if (unlikely(vmx->emulation_required)) {
+
+		/* We don't emulate invalid state of a nested guest */
+		vmx->fail = is_guest_mode(vcpu);
+
+		vmx->exit_reason.full = EXIT_REASON_INVALID_STATE;
+		vmx->exit_reason.failed_vmentry = 1;
+		kvm_register_mark_available(vcpu, VCPU_EXREG_EXIT_INFO_1);
+		vmx->exit_qualification = ENTRY_FAIL_DEFAULT;
+		kvm_register_mark_available(vcpu, VCPU_EXREG_EXIT_INFO_2);
+		vmx->exit_intr_info = 0;
 		return EXIT_FASTPATH_NONE;
+	}
 
 	trace_kvm_entry(vcpu);
 

arch/x86/kvm/vmx/vmx.h

···
 	 * only loaded into hardware when necessary, e.g. SYSCALL #UDs outside
 	 * of 64-bit mode or if EFER.SCE=1, thus the SYSCALL MSRs don't need to
 	 * be loaded into hardware if those conditions aren't met.
-	 * nr_active_uret_msrs tracks the number of MSRs that need to be loaded
-	 * into hardware when running the guest.  guest_uret_msrs[] is resorted
-	 * whenever the number of "active" uret MSRs is modified.
 	 */
 	struct vmx_uret_msr   guest_uret_msrs[MAX_NR_USER_RETURN_MSRS];
-	int                   nr_active_uret_msrs;
 	bool                  guest_uret_msrs_loaded;
 #ifdef CONFIG_X86_64
 	u64		      msr_host_kernel_gs_base;
···
 void vmx_set_host_fs_gs(struct vmcs_host_state *host, u16 fs_sel, u16 gs_sel,
 			unsigned long fs_base, unsigned long gs_base);
 int vmx_get_cpl(struct kvm_vcpu *vcpu);
+bool vmx_emulation_required(struct kvm_vcpu *vcpu);
 unsigned long vmx_get_rflags(struct kvm_vcpu *vcpu);
 void vmx_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags);
 u32 vmx_get_interrupt_shadow(struct kvm_vcpu *vcpu);

arch/x86/kvm/x86.c

···
 	MSR_ARCH_PERFMON_EVENTSEL0 + 12, MSR_ARCH_PERFMON_EVENTSEL0 + 13,
 	MSR_ARCH_PERFMON_EVENTSEL0 + 14, MSR_ARCH_PERFMON_EVENTSEL0 + 15,
 	MSR_ARCH_PERFMON_EVENTSEL0 + 16, MSR_ARCH_PERFMON_EVENTSEL0 + 17,
+
+	MSR_K7_EVNTSEL0, MSR_K7_EVNTSEL1, MSR_K7_EVNTSEL2, MSR_K7_EVNTSEL3,
+	MSR_K7_PERFCTR0, MSR_K7_PERFCTR1, MSR_K7_PERFCTR2, MSR_K7_PERFCTR3,
+	MSR_F15H_PERF_CTL0, MSR_F15H_PERF_CTL1, MSR_F15H_PERF_CTL2,
+	MSR_F15H_PERF_CTL3, MSR_F15H_PERF_CTL4, MSR_F15H_PERF_CTL5,
+	MSR_F15H_PERF_CTR0, MSR_F15H_PERF_CTR1, MSR_F15H_PERF_CTR2,
+	MSR_F15H_PERF_CTR3, MSR_F15H_PERF_CTR4, MSR_F15H_PERF_CTR5,
 };
 
 static u32 msrs_to_save[ARRAY_SIZE(msrs_to_save_all)];
···
 				       offsetof(struct compat_vcpu_info, time));
 	if (vcpu->xen.vcpu_time_info_set)
 		kvm_setup_pvclock_page(v, &vcpu->xen.vcpu_time_info_cache, 0);
-	if (v == kvm_get_vcpu(v->kvm, 0))
+	if (!v->vcpu_idx)
 		kvm_hv_setup_tsc_page(v->kvm, &vcpu->hv_clock);
 	return 0;
 }
···
 
 		/* Process a latched INIT or SMI, if any.  */
 		kvm_make_request(KVM_REQ_EVENT, vcpu);
+
+		/*
+		 * Even if KVM_SET_SREGS2 loaded PDPTRs out of band,
+		 * on SMM exit we still need to reload them from
+		 * guest memory
+		 */
+		vcpu->arch.pdptrs_from_userspace = false;
 	}
 
 	kvm_mmu_reset_context(vcpu);
···
 	int r;
 
 	vcpu->arch.last_vmentry_cpu = -1;
+	vcpu->arch.regs_avail = ~0;
+	vcpu->arch.regs_dirty = ~0;
 
 	if (!irqchip_in_kernel(vcpu->kvm) || kvm_vcpu_is_reset_bsp(vcpu))
 		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
···
 
 	kvm_set_rflags(vcpu, X86_EFLAGS_FIXED);
 	kvm_rip_write(vcpu, 0xfff0);
+
+	vcpu->arch.cr3 = 0;
+	kvm_register_mark_dirty(vcpu, VCPU_EXREG_CR3);
 
 	/*
 	 * CR0.CD/NW are set on RESET, preserved on INIT.  Note, some versions
···
 
 int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 {
+	int ret;
+
 	if (type)
 		return -EINVAL;
+
+	ret = kvm_page_track_init(kvm);
+	if (ret)
+		return ret;
 
 	INIT_HLIST_HEAD(&kvm->arch.mask_notifier_list);
 	INIT_LIST_HEAD(&kvm->arch.active_mmu_pages);
···
 
 	kvm_apicv_init(kvm);
 	kvm_hv_init_vm(kvm);
-	kvm_page_track_init(kvm);
 	kvm_mmu_init_vm(kvm);
 	kvm_xen_init_vm(kvm);
 

drivers/perf/arm_pmu.c

···
 		pmu->name, pmu->num_events,
 		has_nmi ? ", using NMIs" : "");
 
+	kvm_host_pmu_init(pmu);
+
 	return 0;
 
 out_destroy:

include/kvm/arm_pmu.h

···
 int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu,
 			    struct kvm_device_attr *attr);
 int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu);
-int kvm_pmu_probe_pmuver(void);
 #else
 struct kvm_pmu {
 };
···
 {
 	return 0;
 }
-
-static inline int kvm_pmu_probe_pmuver(void) { return 0xf; }
 
 #endif
 

include/linux/kvm_host.h

···
 	unsigned long mmu_notifier_range_start;
 	unsigned long mmu_notifier_range_end;
 #endif
-	long tlbs_dirty;
 	struct list_head devices;
 	u64 manual_dirty_log_protect;
 	struct dentry *debugfs_dentry;
···
 		if (vcpu->vcpu_id == id)
 			return vcpu;
 	return NULL;
-}
-
-static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
-{
-	return vcpu->vcpu_idx;
 }
 
 #define kvm_for_each_memslot(memslot, slots)				\

include/linux/perf/arm_pmu.h

···
 static inline int arm_pmu_acpi_probe(armpmu_init_fn init_fn) { return 0; }
 #endif
 
+#ifdef CONFIG_KVM
+void kvm_host_pmu_init(struct arm_pmu *pmu);
+#else
+#define kvm_host_pmu_init(x)	do { } while(0)
+#endif
+
 /* Internal functions only for core arm_pmu code */
 struct arm_pmu *armpmu_alloc(void);
 struct arm_pmu *armpmu_alloc_atomic(void);

tools/testing/selftests/kvm/.gitignore

···
 /x86_64/smm_test
 /x86_64/state_test
 /x86_64/svm_vmcall_test
+/x86_64/svm_int_ctl_test
 /x86_64/sync_regs_test
 /x86_64/tsc_msrs_test
 /x86_64/userspace_msr_exit_test

tools/testing/selftests/kvm/Makefile

···
 TEST_GEN_PROGS_x86_64 += x86_64/state_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_preemption_timer_test
 TEST_GEN_PROGS_x86_64 += x86_64/svm_vmcall_test
+TEST_GEN_PROGS_x86_64 += x86_64/svm_int_ctl_test
 TEST_GEN_PROGS_x86_64 += x86_64/sync_regs_test
 TEST_GEN_PROGS_x86_64 += x86_64/userspace_msr_exit_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_apic_access_test

tools/testing/selftests/kvm/access_tracking_perf_test.c

···
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n\n");
-	backing_src_help();
+	backing_src_help("-s");
 	puts("");
 	exit(0);
 }
···
 int main(int argc, char *argv[])
 {
 	struct test_params params = {
-		.backing_src = VM_MEM_SRC_ANONYMOUS,
+		.backing_src = DEFAULT_VM_MEM_SRC,
 		.vcpu_memory_bytes = DEFAULT_PER_VCPU_MEM_SIZE,
 		.vcpus = 1,
 	};

tools/testing/selftests/kvm/demand_paging_test.c

···
 			return NULL;
 		}
 
-		if (!pollfd[0].revents & POLLIN)
+		if (!(pollfd[0].revents & POLLIN))
 			continue;
 
 		r = read(uffd, &msg, sizeof(msg));
···
 {
 	puts("");
 	printf("usage: %s [-h] [-m vm_mode] [-u uffd_mode] [-d uffd_delay_usec]\n"
-	       "          [-b memory] [-t type] [-v vcpus] [-o]\n", name);
+	       "          [-b memory] [-s type] [-v vcpus] [-o]\n", name);
 	guest_modes_help();
 	printf(" -u: use userfaultfd to handle vCPU page faults. Mode is a\n"
 	       "     UFFD registration mode: 'MISSING' or 'MINOR'.\n");
···
 	printf(" -b: specify the size of the memory region which should be\n"
 	       "     demand paged by each vCPU. e.g. 10M or 3G.\n"
 	       "     Default: 1G\n");
-	printf(" -t: The type of backing memory to use. Default: anonymous\n");
-	backing_src_help();
+	backing_src_help("-s");
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
···
 {
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
 	struct test_params p = {
-		.src_type = VM_MEM_SRC_ANONYMOUS,
+		.src_type = DEFAULT_VM_MEM_SRC,
 		.partition_vcpu_memory_access = true,
 	};
 	int opt;
 
 	guest_modes_append_default();
 
-	while ((opt = getopt(argc, argv, "hm:u:d:b:t:v:o")) != -1) {
+	while ((opt = getopt(argc, argv, "hm:u:d:b:s:v:o")) != -1) {
 		switch (opt) {
 		case 'm':
 			guest_modes_cmdline(optarg);
···
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
 			break;
-		case 't':
+		case 's':
 			p.src_type = parse_backing_src_type(optarg);
 			break;
 		case 'v':
···
 
 	if (p.uffd_mode == UFFDIO_REGISTER_MODE_MINOR &&
 	    !backing_src_is_shared(p.src_type)) {
-		TEST_FAIL("userfaultfd MINOR mode requires shared memory; pick a different -t");
+		TEST_FAIL("userfaultfd MINOR mode requires shared memory; pick a different -s");
 	}
 
 	for_each_guest_mode(run_test, &p);

tools/testing/selftests/kvm/dirty_log_perf_test.c

···
 	toggle_dirty_logging(vm, slots, false);
 }
 
-static void get_dirty_log(struct kvm_vm *vm, int slots, unsigned long *bitmap,
-			  uint64_t nr_pages)
+static void get_dirty_log(struct kvm_vm *vm, unsigned long *bitmaps[], int slots)
 {
-	uint64_t slot_pages = nr_pages / slots;
 	int i;
 
 	for (i = 0; i < slots; i++) {
 		int slot = PERF_TEST_MEM_SLOT_INDEX + i;
-		unsigned long *slot_bitmap = bitmap + i * slot_pages;
 
-		kvm_vm_get_dirty_log(vm, slot, slot_bitmap);
+		kvm_vm_get_dirty_log(vm, slot, bitmaps[i]);
 	}
 }
 
-static void clear_dirty_log(struct kvm_vm *vm, int slots, unsigned long *bitmap,
-			    uint64_t nr_pages)
+static void clear_dirty_log(struct kvm_vm *vm, unsigned long *bitmaps[],
+			    int slots, uint64_t pages_per_slot)
 {
-	uint64_t slot_pages = nr_pages / slots;
 	int i;
 
 	for (i = 0; i < slots; i++) {
 		int slot = PERF_TEST_MEM_SLOT_INDEX + i;
-		unsigned long *slot_bitmap = bitmap + i * slot_pages;
 
-		kvm_vm_clear_dirty_log(vm, slot, slot_bitmap, 0, slot_pages);
+		kvm_vm_clear_dirty_log(vm, slot, bitmaps[i], 0, pages_per_slot);
 	}
+}
+
+static unsigned long **alloc_bitmaps(int slots, uint64_t pages_per_slot)
+{
+	unsigned long **bitmaps;
+	int i;
+
+	bitmaps = malloc(slots * sizeof(bitmaps[0]));
+	TEST_ASSERT(bitmaps, "Failed to allocate bitmaps array.");
+
+	for (i = 0; i < slots; i++) {
+		bitmaps[i] = bitmap_zalloc(pages_per_slot);
+		TEST_ASSERT(bitmaps[i], "Failed to allocate slot bitmap.");
+	}
+
+	return bitmaps;
+}
+
+static void free_bitmaps(unsigned long *bitmaps[], int slots)
+{
+	int i;
+
+	for (i = 0; i < slots; i++)
+		free(bitmaps[i]);
+
+	free(bitmaps);
 }
 
 static void run_test(enum vm_guest_mode mode, void *arg)
···
 	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	struct kvm_vm *vm;
-	unsigned long *bmap;
+	unsigned long **bitmaps;
 	uint64_t guest_num_pages;
 	uint64_t host_num_pages;
+	uint64_t pages_per_slot;
 	int vcpu_id;
 	struct timespec start;
 	struct timespec ts_diff;
···
 	guest_num_pages = (nr_vcpus * guest_percpu_mem_size) >> vm_get_page_shift(vm);
 	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
 	host_num_pages = vm_num_host_pages(mode, guest_num_pages);
-	bmap = bitmap_zalloc(host_num_pages);
+	pages_per_slot = host_num_pages / p->slots;
+
+	bitmaps = alloc_bitmaps(p->slots, pages_per_slot);
 
 	if (dirty_log_manual_caps) {
 		cap.cap = KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2;
···
 			iteration, ts_diff.tv_sec, ts_diff.tv_nsec);
 
 		clock_gettime(CLOCK_MONOTONIC, &start);
-		get_dirty_log(vm, p->slots, bmap, host_num_pages);
+		get_dirty_log(vm, bitmaps, p->slots);
 		ts_diff = timespec_elapsed(start);
 		get_dirty_log_total = timespec_add(get_dirty_log_total,
 						   ts_diff);
···
 
 		if (dirty_log_manual_caps) {
 			clock_gettime(CLOCK_MONOTONIC, &start);
-			clear_dirty_log(vm, p->slots, bmap, host_num_pages);
+			clear_dirty_log(vm, bitmaps, p->slots, pages_per_slot);
 			ts_diff = timespec_elapsed(start);
 			clear_dirty_log_total = timespec_add(clear_dirty_log_total,
 							     ts_diff);
···
 			clear_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 	}
 
-	free(bmap);
+	free_bitmaps(bitmaps, p->slots);
 	free(vcpu_threads);
 	perf_test_destroy_vm(vm);
 }
···
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n\n");
+	backing_src_help("-s");
 	printf(" -x: Split the memory region into this number of memslots.\n"
-	       "     (default: 1)");
-	backing_src_help();
+	       "     (default: 1)\n");
 	puts("");
 	exit(0);
 }
···
 		.iterations = TEST_HOST_LOOP_N,
 		.wr_fract = 1,
 		.partition_vcpu_memory_access = true,
-		.backing_src = VM_MEM_SRC_ANONYMOUS,
+		.backing_src = DEFAULT_VM_MEM_SRC,
 		.slots = 1,
 	};
 	int opt;

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

···
 	NUM_SRC_TYPES,
 };
 
+#define DEFAULT_VM_MEM_SRC VM_MEM_SRC_ANONYMOUS
+
 struct vm_mem_backing_src_alias {
 	const char *name;
 	uint32_t flag;
···
 size_t get_def_hugetlb_pagesz(void);
 const struct vm_mem_backing_src_alias *vm_mem_backing_src_alias(uint32_t i);
 size_t get_backing_src_pagesz(uint32_t i);
-void backing_src_help(void);
+void backing_src_help(const char *flag);
 enum vm_mem_backing_src_type parse_backing_src_type(const char *type_name);
 long get_run_delay(void);
 

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

···
 	}
 }
 
-typedef unsigned long v1di __attribute__ ((vector_size (8)));
+#define GET_XMM(__xmm)							\
+({									\
+	unsigned long __val;						\
+	asm volatile("movq %%"#__xmm", %0" : "=r"(__val) : : #__xmm);	\
+	__val;								\
+})
+
 static inline unsigned long get_xmm(int n)
 {
 	assert(n >= 0 && n <= 7);
 
-	register v1di xmm0 __asm__("%xmm0");
-	register v1di xmm1 __asm__("%xmm1");
-	register v1di xmm2 __asm__("%xmm2");
-	register v1di xmm3 __asm__("%xmm3");
-	register v1di xmm4 __asm__("%xmm4");
-	register v1di xmm5 __asm__("%xmm5");
-	register v1di xmm6 __asm__("%xmm6");
-	register v1di xmm7 __asm__("%xmm7");
 	switch (n) {
 	case 0:
-		return (unsigned long)xmm0;
+		return GET_XMM(xmm0);
 	case 1:
-		return (unsigned long)xmm1;
+		return GET_XMM(xmm1);
 	case 2:
-		return (unsigned long)xmm2;
+		return GET_XMM(xmm2);
 	case 3:
-		return (unsigned long)xmm3;
+		return GET_XMM(xmm3);
 	case 4:
-		return (unsigned long)xmm4;
+		return GET_XMM(xmm4);
 	case 5:
-		return (unsigned long)xmm5;
+		return GET_XMM(xmm5);
 	case 6:
-		return (unsigned long)xmm6;
+		return GET_XMM(xmm6);
 	case 7:
-		return (unsigned long)xmm7;
+		return GET_XMM(xmm7);
 	}
+
+	/* never reached */
 	return 0;
 }
 

tools/testing/selftests/kvm/kvm_page_table_test.c

···
 	       "     (default: 1G)\n");
 	printf(" -v: specify the number of vCPUs to run\n"
 	       "     (default: 1)\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n"
-	       "     (default: anonymous)\n\n");
-	backing_src_help();
+	backing_src_help("-s");
 	puts("");
 }
 
···
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
 	struct test_params p = {
 		.test_mem_size = DEFAULT_TEST_MEM_SIZE,
-		.src_type = VM_MEM_SRC_ANONYMOUS,
+		.src_type = DEFAULT_VM_MEM_SRC,
 	};
 	int opt;
 

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

···
 	}
 }
 
-void backing_src_help(void)
+static void print_available_backing_src_types(const char *prefix)
 {
 	int i;
 
-	printf("Available backing src types:\n");
+	printf("%sAvailable backing src types:\n", prefix);
+
 	for (i = 0; i < NUM_SRC_TYPES; i++)
-		printf("\t%s\n", vm_mem_backing_src_alias(i)->name);
+		printf("%s    %s\n", prefix, vm_mem_backing_src_alias(i)->name);
+}
+
+void backing_src_help(const char *flag)
+{
+	printf(" %s: specify the type of memory that should be used to\n"
+	       "     back the guest data region. (default: %s)\n",
+	       flag, vm_mem_backing_src_alias(DEFAULT_VM_MEM_SRC)->name);
+	print_available_backing_src_types("     ");
 }
 
 enum vm_mem_backing_src_type parse_backing_src_type(const char *type_name)
···
 		if (!strcmp(type_name, vm_mem_backing_src_alias(i)->name))
 			return i;
 
-	backing_src_help();
+	print_available_backing_src_types("");
 	TEST_FAIL("Unknown backing src type: %s", type_name);
 	return -1;
 }

tools/testing/selftests/kvm/rseq_test.c

···
 	 * CPU affinity.
 	 */
 	vm = vm_create_default(VCPU_ID, 0, guest_code);
+	ucall_init(vm, NULL);
 
 	pthread_create(&migration_thread, NULL, migration_worker, 0);
 

tools/testing/selftests/kvm/steal_time.c

···
 	uint64_t st_time;
 };
 
-static int64_t smccc(uint32_t func, uint32_t arg)
+static int64_t smccc(uint32_t func, uint64_t arg)
 {
 	unsigned long ret;
 
 	asm volatile(
-		"mov	x0, %1\n"
+		"mov	w0, %w1\n"
 		"mov	x1, %2\n"
 		"hvc	#0\n"
 		"mov	%0, x0\n"

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

···
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * svm_int_ctl_test
+ *
+ * Copyright (C) 2021, Red Hat, Inc.
+ *
+ * Nested SVM testing: test simultaneous use of V_IRQ from L1 and L0.
+ */
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "processor.h"
+#include "svm_util.h"
+#include "apic.h"
+
+#define VCPU_ID		0
+
+static struct kvm_vm *vm;
+
+bool vintr_irq_called;
+bool intr_irq_called;
+
+#define VINTR_IRQ_NUMBER 0x20
+#define INTR_IRQ_NUMBER 0x30
+
+static void vintr_irq_handler(struct ex_regs *regs)
+{
+	vintr_irq_called = true;
+}
+
+static void intr_irq_handler(struct ex_regs *regs)
+{
+	x2apic_write_reg(APIC_EOI, 0x00);
+	intr_irq_called = true;
+}
+
+static void l2_guest_code(struct svm_test_data *svm)
+{
+	/* This code raises interrupt INTR_IRQ_NUMBER in the L1's LAPIC,
+	 * and since L1 didn't enable virtual interrupt masking,
+	 * L2 should receive it and not L1.
+	 *
+	 * L2 also has virtual interrupt 'VINTR_IRQ_NUMBER' pending in V_IRQ
+	 * so it should also receive it after the following 'sti'.
+	 */
+	x2apic_write_reg(APIC_ICR,
+		APIC_DEST_SELF | APIC_INT_ASSERT | INTR_IRQ_NUMBER);
+
+	__asm__ __volatile__(
+		"sti\n"
+		"nop\n"
+	);
+
+	GUEST_ASSERT(vintr_irq_called);
+	GUEST_ASSERT(intr_irq_called);
+
+	__asm__ __volatile__(
+		"vmcall\n"
+	);
+}
+
+static void l1_guest_code(struct svm_test_data *svm)
+{
+	#define L2_GUEST_STACK_SIZE 64
+	unsigned long l2_guest_stack[L2_GUEST_STACK_SIZE];
+	struct vmcb *vmcb = svm->vmcb;
+
+	x2apic_enable();
+
+	/* Prepare for L2 execution. */
+	generic_svm_setup(svm, l2_guest_code,
+			  &l2_guest_stack[L2_GUEST_STACK_SIZE]);
+
+	/* No virtual interrupt masking */
+	vmcb->control.int_ctl &= ~V_INTR_MASKING_MASK;
+
+	/* No intercepts for real and virtual interrupts */
+	vmcb->control.intercept &= ~(1ULL << INTERCEPT_INTR | INTERCEPT_VINTR);
+
+	/* Make a virtual interrupt VINTR_IRQ_NUMBER pending */
+	vmcb->control.int_ctl |= V_IRQ_MASK | (0x1 << V_INTR_PRIO_SHIFT);
+	vmcb->control.int_vector = VINTR_IRQ_NUMBER;
+
+	run_guest(vmcb, svm->vmcb_gpa);
+	GUEST_ASSERT(vmcb->control.exit_code == SVM_EXIT_VMMCALL);
+	GUEST_DONE();
+}
+
+int main(int argc, char *argv[])
+{
+	vm_vaddr_t svm_gva;
+
+	nested_svm_check_supported();
+
+	vm = vm_create_default(VCPU_ID, 0, (void *) l1_guest_code);
+
+	vm_init_descriptor_tables(vm);
+	vcpu_init_descriptor_tables(vm, VCPU_ID);
+
+	vm_install_exception_handler(vm, VINTR_IRQ_NUMBER, vintr_irq_handler);
+	vm_install_exception_handler(vm, INTR_IRQ_NUMBER, intr_irq_handler);
+
+	vcpu_alloc_svm(vm, &svm_gva);
+	vcpu_args_set(vm, VCPU_ID, 1, svm_gva);
+
+	struct kvm_run *run = vcpu_state(vm, VCPU_ID);
+	struct ucall uc;
+
+	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));
+
+	switch (get_ucall(vm, VCPU_ID, &uc)) {
+	case UCALL_ABORT:
+		TEST_FAIL("%s", (const char *)uc.args[0]);
+		break;
+		/* NOT REACHED */
+	case UCALL_DONE:
+		goto done;
+	default:
+		TEST_FAIL("Unknown ucall 0x%lx.", uc.cmd);
+	}
+done:
+	kvm_vm_free(vm);
+	return 0;
+}

virt/kvm/kvm_main.c

···
 {
 }
 
-static inline bool kvm_kick_many_cpus(const struct cpumask *cpus, bool wait)
+static inline bool kvm_kick_many_cpus(cpumask_var_t tmp, bool wait)
 {
-	if (unlikely(!cpus))
+	const struct cpumask *cpus;
+
+	if (likely(cpumask_available(tmp)))
+		cpus = tmp;
+	else
 		cpus = cpu_online_mask;
 
 	if (cpumask_empty(cpus))
···
 			continue;
 
 		kvm_make_request(req, vcpu);
-		cpu = vcpu->cpu;
 
 		if (!(req & KVM_REQUEST_NO_WAKEUP) && kvm_vcpu_wake_up(vcpu))
 			continue;
 
-		if (tmp != NULL && cpu != -1 && cpu != me &&
-		    kvm_request_needs_ipi(vcpu, req))
-			__cpumask_set_cpu(cpu, tmp);
+		/*
+		 * tmp can be "unavailable" if cpumasks are allocated off stack
+		 * as allocation of the mask is deliberately not fatal and is
+		 * handled by falling back to kicking all online CPUs.
+		 */
+		if (!cpumask_available(tmp))
+			continue;
+
+		/*
+		 * Note, the vCPU could get migrated to a different pCPU at any
+		 * point after kvm_request_needs_ipi(), which could result in
+		 * sending an IPI to the previous pCPU.  But, that's ok because
+		 * the purpose of the IPI is to ensure the vCPU returns to
+		 * OUTSIDE_GUEST_MODE, which is satisfied if the vCPU migrates.
+		 * Entering READING_SHADOW_PAGE_TABLES after this point is also
+		 * ok, as the requirement is only that KVM wait for vCPUs that
+		 * were reading SPTEs _before_ any changes were finalized.  See
+		 * kvm_vcpu_kick() for more details on handling requests.
+		 */
+		if (kvm_request_needs_ipi(vcpu, req)) {
+			cpu = READ_ONCE(vcpu->cpu);
+			if (cpu != -1 && cpu != me)
+				__cpumask_set_cpu(cpu, tmp);
+		}
 	}
 
 	called = kvm_kick_many_cpus(tmp, !!(req & KVM_REQUEST_WAIT));
···
 #ifndef CONFIG_HAVE_KVM_ARCH_TLB_FLUSH_ALL
 void kvm_flush_remote_tlbs(struct kvm *kvm)
 {
-	/*
-	 * Read tlbs_dirty before setting KVM_REQ_TLB_FLUSH in
-	 * kvm_make_all_cpus_request.
-	 */
-	long dirty_count = smp_load_acquire(&kvm->tlbs_dirty);
-
 	++kvm->stat.generic.remote_tlb_flush_requests;
+
 	/*
 	 * We want to publish modifications to the page tables before reading
 	 * mode. Pairs with a memory barrier in arch-specific code.
···
 	if (!kvm_arch_flush_remote_tlb(kvm)
 	    || kvm_make_all_cpus_request(kvm, KVM_REQ_TLB_FLUSH))
 		++kvm->stat.generic.remote_tlb_flush;
-	cmpxchg(&kvm->tlbs_dirty, dirty_count, 0);
 }
 EXPORT_SYMBOL_GPL(kvm_flush_remote_tlbs);
 #endif
···
 		}
 	}
 
-	if (range->flush_on_ret && (ret || kvm->tlbs_dirty))
+	if (range->flush_on_ret && ret)
 		kvm_flush_remote_tlbs(kvm);
 
 	if (locked)
···
 
 static void shrink_halt_poll_ns(struct kvm_vcpu *vcpu)
 {
-	unsigned int old, val, shrink;
+	unsigned int old, val, shrink, grow_start;
 
 	old = val = vcpu->halt_poll_ns;
 	shrink = READ_ONCE(halt_poll_ns_shrink);
+	grow_start = READ_ONCE(halt_poll_ns_grow_start);
 	if (shrink == 0)
 		val = 0;
 	else
 		val /= shrink;
+
+	if (val < grow_start)
+		val = 0;
 
 	vcpu->halt_poll_ns = val;
 	trace_kvm_halt_poll_ns_shrink(vcpu->vcpu_id, val, old);
···
  */
 void kvm_vcpu_kick(struct kvm_vcpu *vcpu)
 {
-	int me;
-	int cpu = vcpu->cpu;
+	int me, cpu;
 
 	if (kvm_vcpu_wake_up(vcpu))
 		return;
 
+	/*
+	 * Note, the vCPU could get migrated to a different pCPU at any point
+	 * after kvm_arch_vcpu_should_kick(), which could result in sending an
+	 * IPI to the previous pCPU.  But, that's ok because the purpose of the
+	 * IPI is to force the vCPU to leave IN_GUEST_MODE, and migrating the
+	 * vCPU also requires it to leave IN_GUEST_MODE.
+	 */
 	me = get_cpu();
-	if (cpu != me && (unsigned)cpu < nr_cpu_ids && cpu_online(cpu))
-		if (kvm_arch_vcpu_should_kick(vcpu))
+	if (kvm_arch_vcpu_should_kick(vcpu)) {
+		cpu = READ_ONCE(vcpu->cpu);
+		if (cpu != me && (unsigned)cpu < nr_cpu_ids && cpu_online(cpu))
 			smp_send_reschedule(cpu);
+	}
 	put_cpu();
 }
 EXPORT_SYMBOL_GPL(kvm_vcpu_kick);