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

Pull kvm fixes from Paolo Bonzini:
 "Bugfixes, many of them reported by syzkaller and mostly predating the
  merge window"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  kvm: svm: Ensure an IBPB on all affected CPUs when freeing a vmcb
  kvm: mmu: Fix race in emulated page table writes
  KVM: nVMX: vmcs12 revision_id is always VMCS12_REVISION even when copied from eVMCS
  KVM: nVMX: Verify eVMCS revision id match supported eVMCS version on eVMCS VMPTRLD
  KVM: nVMX/nSVM: Fix bug which sets vcpu->arch.tsc_offset to L1 tsc_offset
  x86/kvm/vmx: fix old-style function declaration
  KVM: x86: fix empty-body warnings
  KVM: VMX: Update shared MSRs to be saved/restored on MSR_EFER.LMA changes
  KVM: x86: Fix kernel info-leak in KVM_HC_CLOCK_PAIRING hypercall
  KVM: nVMX: Fix kernel info-leak when enabling KVM_CAP_HYPERV_ENLIGHTENED_VMCS more than once
  svm: Add mutex_lock to protect apic_access_page_done on AMD systems
  KVM: X86: Fix scan ioapic use-before-initialization
  KVM: LAPIC: Fix pv ipis use-before-initialization
  KVM: VMX: re-add ple_gap module parameter
  KVM: PPC: Book3S HV: Fix handling for interrupted H_ENTER_NESTED

arch/powerpc/kvm/book3s_hv.c

@@ -983,6 +983,7 @@ int kvmppc_pseries_do_hcall(struct kvm_vcpu *vcpu)
 		ret = kvmhv_enter_nested_guest(vcpu);
 		if (ret == H_INTERRUPT) {
 			kvmppc_set_gpr(vcpu, 3, 0);
+			vcpu->arch.hcall_needed = 0;
 			return -EINTR;
 		}
 		break;

arch/x86/include/asm/kvm_host.h

@@ -1094,7 +1094,8 @@ struct kvm_x86_ops {
 	bool (*has_wbinvd_exit)(void);
 
 	u64 (*read_l1_tsc_offset)(struct kvm_vcpu *vcpu);
-	void (*write_tsc_offset)(struct kvm_vcpu *vcpu, u64 offset);
+	/* Returns actual tsc_offset set in active VMCS */
+	u64 (*write_l1_tsc_offset)(struct kvm_vcpu *vcpu, u64 offset);
 
 	void (*get_exit_info)(struct kvm_vcpu *vcpu, u64 *info1, u64 *info2);
 

arch/x86/kvm/lapic.c

@@ -55,7 +55,7 @@
 #define PRIo64 "o"
 
 /* #define apic_debug(fmt,arg...) printk(KERN_WARNING fmt,##arg) */
-#define apic_debug(fmt, arg...)
+#define apic_debug(fmt, arg...) do {} while (0)
 
 /* 14 is the version for Xeon and Pentium 8.4.8*/
 #define APIC_VERSION			(0x14UL | ((KVM_APIC_LVT_NUM - 1) << 16))
@@ -576,6 +576,11 @@ int kvm_pv_send_ipi(struct kvm *kvm, unsigned long ipi_bitmap_low,
 	rcu_read_lock();
 	map = rcu_dereference(kvm->arch.apic_map);
 
+	if (unlikely(!map)) {
+		count = -EOPNOTSUPP;
+		goto out;
+	}
+
 	if (min > map->max_apic_id)
 		goto out;
 	/* Bits above cluster_size are masked in the caller.  */

arch/x86/kvm/mmu.c

@@ -5074,9 +5074,9 @@ static bool need_remote_flush(u64 old, u64 new)
 }
 
 static u64 mmu_pte_write_fetch_gpte(struct kvm_vcpu *vcpu, gpa_t *gpa,
-				    const u8 *new, int *bytes)
+				    int *bytes)
 {
-	u64 gentry;
+	u64 gentry = 0;
 	int r;
 
 	/*
@@ -5088,22 +5088,12 @@ static u64 mmu_pte_write_fetch_gpte(struct kvm_vcpu *vcpu, gpa_t *gpa,
 		/* Handle a 32-bit guest writing two halves of a 64-bit gpte */
 		*gpa &= ~(gpa_t)7;
 		*bytes = 8;
-		r = kvm_vcpu_read_guest(vcpu, *gpa, &gentry, 8);
-		if (r)
-			gentry = 0;
-		new = (const u8 *)&gentry;
 	}
 
-	switch (*bytes) {
-	case 4:
-		gentry = *(const u32 *)new;
-		break;
-	case 8:
-		gentry = *(const u64 *)new;
-		break;
-	default:
-		gentry = 0;
-		break;
+	if (*bytes == 4 || *bytes == 8) {
+		r = kvm_vcpu_read_guest_atomic(vcpu, *gpa, &gentry, *bytes);
+		if (r)
+			gentry = 0;
 	}
 
 	return gentry;
@@ -5207,8 +5197,6 @@ static void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 
 	pgprintk("%s: gpa %llx bytes %d\n", __func__, gpa, bytes);
 
-	gentry = mmu_pte_write_fetch_gpte(vcpu, &gpa, new, &bytes);
-
 	/*
 	 * No need to care whether allocation memory is successful
 	 * or not since pte prefetch is skiped if it does not have
@@ -5217,6 +5205,9 @@ static void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	mmu_topup_memory_caches(vcpu);
 
 	spin_lock(&vcpu->kvm->mmu_lock);
+
+	gentry = mmu_pte_write_fetch_gpte(vcpu, &gpa, &bytes);
+
 	++vcpu->kvm->stat.mmu_pte_write;
 	kvm_mmu_audit(vcpu, AUDIT_PRE_PTE_WRITE);
 

arch/x86/kvm/svm.c

@@ -1446,7 +1446,7 @@ static u64 svm_read_l1_tsc_offset(struct kvm_vcpu *vcpu)
 	return vcpu->arch.tsc_offset;
 }
 
-static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
+static u64 svm_write_l1_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 	u64 g_tsc_offset = 0;
@@ -1464,6 +1464,7 @@ static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 	svm->vmcb->control.tsc_offset = offset + g_tsc_offset;
 
 	mark_dirty(svm->vmcb, VMCB_INTERCEPTS);
+	return svm->vmcb->control.tsc_offset;
 }
 
 static void avic_init_vmcb(struct vcpu_svm *svm)
@@ -1664,20 +1665,23 @@ static u64 *avic_get_physical_id_entry(struct kvm_vcpu *vcpu,
 static int avic_init_access_page(struct kvm_vcpu *vcpu)
 {
 	struct kvm *kvm = vcpu->kvm;
-	int ret;
+	int ret = 0;
 
+	mutex_lock(&kvm->slots_lock);
 	if (kvm->arch.apic_access_page_done)
-		return 0;
+		goto out;
 
-	ret = x86_set_memory_region(kvm,
-				    APIC_ACCESS_PAGE_PRIVATE_MEMSLOT,
-				    APIC_DEFAULT_PHYS_BASE,
-				    PAGE_SIZE);
+	ret = __x86_set_memory_region(kvm,
+				      APIC_ACCESS_PAGE_PRIVATE_MEMSLOT,
+				      APIC_DEFAULT_PHYS_BASE,
+				      PAGE_SIZE);
 	if (ret)
-		return ret;
+		goto out;
 
 	kvm->arch.apic_access_page_done = true;
-	return 0;
+out:
+	mutex_unlock(&kvm->slots_lock);
+	return ret;
 }
 
 static int avic_init_backing_page(struct kvm_vcpu *vcpu)
@@ -2189,21 +2193,31 @@ static struct kvm_vcpu *svm_create_vcpu(struct kvm *kvm, unsigned int id)
 	return ERR_PTR(err);
 }
 
+static void svm_clear_current_vmcb(struct vmcb *vmcb)
+{
+	int i;
+
+	for_each_online_cpu(i)
+		cmpxchg(&per_cpu(svm_data, i)->current_vmcb, vmcb, NULL);
+}
+
 static void svm_free_vcpu(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 
+	/*
+	 * The vmcb page can be recycled, causing a false negative in
+	 * svm_vcpu_load(). So, ensure that no logical CPU has this
+	 * vmcb page recorded as its current vmcb.
+	 */
+	svm_clear_current_vmcb(svm->vmcb);
+
 	__free_page(pfn_to_page(__sme_clr(svm->vmcb_pa) >> PAGE_SHIFT));
 	__free_pages(virt_to_page(svm->msrpm), MSRPM_ALLOC_ORDER);
 	__free_page(virt_to_page(svm->nested.hsave));
 	__free_pages(virt_to_page(svm->nested.msrpm), MSRPM_ALLOC_ORDER);
 	kvm_vcpu_uninit(vcpu);
 	kmem_cache_free(kvm_vcpu_cache, svm);
-	/*
-	 * The vmcb page can be recycled, causing a false negative in
-	 * svm_vcpu_load(). So do a full IBPB now.
-	 */
-	indirect_branch_prediction_barrier();
 }
 
 static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
@@ -7149,7 +7163,7 @@ static struct kvm_x86_ops svm_x86_ops __ro_after_init = {
 	.has_wbinvd_exit = svm_has_wbinvd_exit,
 
 	.read_l1_tsc_offset = svm_read_l1_tsc_offset,
-	.write_tsc_offset = svm_write_tsc_offset,
+	.write_l1_tsc_offset = svm_write_l1_tsc_offset,
 
 	.set_tdp_cr3 = set_tdp_cr3,
 

arch/x86/kvm/vmx.c

@@ -174,6 +174,7 @@ module_param_named(preemption_timer, enable_preemption_timer, bool, S_IRUGO);
  * refer SDM volume 3b section 21.6.13 & 22.1.3.
  */
 static unsigned int ple_gap = KVM_DEFAULT_PLE_GAP;
+module_param(ple_gap, uint, 0444);
 
 static unsigned int ple_window = KVM_VMX_DEFAULT_PLE_WINDOW;
 module_param(ple_window, uint, 0444);
@@ -984,6 +985,7 @@ struct vcpu_vmx {
 	struct shared_msr_entry *guest_msrs;
 	int                   nmsrs;
 	int                   save_nmsrs;
+	bool                  guest_msrs_dirty;
 	unsigned long	      host_idt_base;
 #ifdef CONFIG_X86_64
 	u64 		      msr_host_kernel_gs_base;
@@ -1306,7 +1308,7 @@ static void vmx_set_nmi_mask(struct kvm_vcpu *vcpu, bool masked);
 static bool nested_vmx_is_page_fault_vmexit(struct vmcs12 *vmcs12,
 					    u16 error_code);
 static void vmx_update_msr_bitmap(struct kvm_vcpu *vcpu);
-static void __always_inline vmx_disable_intercept_for_msr(unsigned long *msr_bitmap,
+static __always_inline void vmx_disable_intercept_for_msr(unsigned long *msr_bitmap,
 							  u32 msr, int type);
 
 static DEFINE_PER_CPU(struct vmcs *, vmxarea);
@@ -1610,12 +1612,6 @@ static int nested_enable_evmcs(struct kvm_vcpu *vcpu,
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
 
-	/* We don't support disabling the feature for simplicity. */
-	if (vmx->nested.enlightened_vmcs_enabled)
-		return 0;
-
-	vmx->nested.enlightened_vmcs_enabled = true;
-
 	/*
 	 * vmcs_version represents the range of supported Enlightened VMCS
 	 * versions: lower 8 bits is the minimal version, higher 8 bits is the
@@ -1625,6 +1621,12 @@ static int nested_enable_evmcs(struct kvm_vcpu *vcpu,
 	if (vmcs_version)
 		*vmcs_version = (KVM_EVMCS_VERSION << 8) | 1;
 
+	/* We don't support disabling the feature for simplicity. */
+	if (vmx->nested.enlightened_vmcs_enabled)
+		return 0;
+
+	vmx->nested.enlightened_vmcs_enabled = true;
+
 	vmx->nested.msrs.pinbased_ctls_high &= ~EVMCS1_UNSUPPORTED_PINCTRL;
 	vmx->nested.msrs.entry_ctls_high &= ~EVMCS1_UNSUPPORTED_VMENTRY_CTRL;
 	vmx->nested.msrs.exit_ctls_high &= ~EVMCS1_UNSUPPORTED_VMEXIT_CTRL;
@@ -2897,6 +2899,20 @@ static void vmx_prepare_switch_to_guest(struct kvm_vcpu *vcpu)
 
 	vmx->req_immediate_exit = false;
 
+	/*
+	 * Note that guest MSRs to be saved/restored can also be changed
+	 * when guest state is loaded. This happens when guest transitions
+	 * to/from long-mode by setting MSR_EFER.LMA.
+	 */
+	if (!vmx->loaded_cpu_state || vmx->guest_msrs_dirty) {
+		vmx->guest_msrs_dirty = false;
+		for (i = 0; i < vmx->save_nmsrs; ++i)
+			kvm_set_shared_msr(vmx->guest_msrs[i].index,
+					   vmx->guest_msrs[i].data,
+					   vmx->guest_msrs[i].mask);
+
+	}
+
 	if (vmx->loaded_cpu_state)
 		return;
 
@@ -2957,11 +2973,6 @@ static void vmx_prepare_switch_to_guest(struct kvm_vcpu *vcpu)
 		vmcs_writel(HOST_GS_BASE, gs_base);
 		host_state->gs_base = gs_base;
 	}
-
-	for (i = 0; i < vmx->save_nmsrs; ++i)
-		kvm_set_shared_msr(vmx->guest_msrs[i].index,
-				   vmx->guest_msrs[i].data,
-				   vmx->guest_msrs[i].mask);
 }
 
 static void vmx_prepare_switch_to_host(struct vcpu_vmx *vmx)
@@ -3436,6 +3447,7 @@ static void setup_msrs(struct vcpu_vmx *vmx)
 		move_msr_up(vmx, index, save_nmsrs++);
 
 	vmx->save_nmsrs = save_nmsrs;
+	vmx->guest_msrs_dirty = true;
 
 	if (cpu_has_vmx_msr_bitmap())
 		vmx_update_msr_bitmap(&vmx->vcpu);
@@ -3452,11 +3464,9 @@ static u64 vmx_read_l1_tsc_offset(struct kvm_vcpu *vcpu)
 	return vcpu->arch.tsc_offset;
 }
 
-/*
- * writes 'offset' into guest's timestamp counter offset register
- */
-static void vmx_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
+static u64 vmx_write_l1_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 {
+	u64 active_offset = offset;
 	if (is_guest_mode(vcpu)) {
 		/*
 		 * We're here if L1 chose not to trap WRMSR to TSC. According
@@ -3464,17 +3474,16 @@ static void vmx_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 		 * set for L2 remains unchanged, and still needs to be added
 		 * to the newly set TSC to get L2's TSC.
 		 */
-		struct vmcs12 *vmcs12;
-		/* recalculate vmcs02.TSC_OFFSET: */
-		vmcs12 = get_vmcs12(vcpu);
-		vmcs_write64(TSC_OFFSET, offset +
-			(nested_cpu_has(vmcs12, CPU_BASED_USE_TSC_OFFSETING) ?
-			 vmcs12->tsc_offset : 0));
+		struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
+		if (nested_cpu_has(vmcs12, CPU_BASED_USE_TSC_OFFSETING))
+			active_offset += vmcs12->tsc_offset;
 	} else {
 		trace_kvm_write_tsc_offset(vcpu->vcpu_id,
 					   vmcs_read64(TSC_OFFSET), offset);
-		vmcs_write64(TSC_OFFSET, offset);
 	}
+
+	vmcs_write64(TSC_OFFSET, active_offset);
+	return active_offset;
 }
 
 /*
@@ -5944,7 +5953,7 @@ static void free_vpid(int vpid)
 	spin_unlock(&vmx_vpid_lock);
 }
 
-static void __always_inline vmx_disable_intercept_for_msr(unsigned long *msr_bitmap,
+static __always_inline void vmx_disable_intercept_for_msr(unsigned long *msr_bitmap,
 							  u32 msr, int type)
 {
 	int f = sizeof(unsigned long);
@@ -5982,7 +5991,7 @@ static void __always_inline vmx_disable_intercept_for_msr(unsigned long *msr_bit
 	}
 }
 
-static void __always_inline vmx_enable_intercept_for_msr(unsigned long *msr_bitmap,
+static __always_inline void vmx_enable_intercept_for_msr(unsigned long *msr_bitmap,
 							 u32 msr, int type)
 {
 	int f = sizeof(unsigned long);
@@ -6020,7 +6029,7 @@ static void __always_inline vmx_enable_intercept_for_msr(unsigned long *msr_bitm
 	}
 }
 
-static void __always_inline vmx_set_intercept_for_msr(unsigned long *msr_bitmap,
+static __always_inline void vmx_set_intercept_for_msr(unsigned long *msr_bitmap,
 			     			      u32 msr, int type, bool value)
 {
 	if (value)
@@ -8664,8 +8673,6 @@ static int copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx)
 	struct vmcs12 *vmcs12 = vmx->nested.cached_vmcs12;
 	struct hv_enlightened_vmcs *evmcs = vmx->nested.hv_evmcs;
 
-	vmcs12->hdr.revision_id = evmcs->revision_id;
-
 	/* HV_VMX_ENLIGHTENED_CLEAN_FIELD_NONE */
 	vmcs12->tpr_threshold = evmcs->tpr_threshold;
 	vmcs12->guest_rip = evmcs->guest_rip;
@@ -9369,7 +9376,30 @@ static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu,
 
 		vmx->nested.hv_evmcs = kmap(vmx->nested.hv_evmcs_page);
 
-		if (vmx->nested.hv_evmcs->revision_id != VMCS12_REVISION) {
+		/*
+		 * Currently, KVM only supports eVMCS version 1
+		 * (== KVM_EVMCS_VERSION) and thus we expect guest to set this
+		 * value to first u32 field of eVMCS which should specify eVMCS
+		 * VersionNumber.
+		 *
+		 * Guest should be aware of supported eVMCS versions by host by
+		 * examining CPUID.0x4000000A.EAX[0:15]. Host userspace VMM is
+		 * expected to set this CPUID leaf according to the value
+		 * returned in vmcs_version from nested_enable_evmcs().
+		 *
+		 * However, it turns out that Microsoft Hyper-V fails to comply
+		 * to their own invented interface: When Hyper-V use eVMCS, it
+		 * just sets first u32 field of eVMCS to revision_id specified
+		 * in MSR_IA32_VMX_BASIC. Instead of used eVMCS version number
+		 * which is one of the supported versions specified in
+		 * CPUID.0x4000000A.EAX[0:15].
+		 *
+		 * To overcome Hyper-V bug, we accept here either a supported
+		 * eVMCS version or VMCS12 revision_id as valid values for first
+		 * u32 field of eVMCS.
+		 */
+		if ((vmx->nested.hv_evmcs->revision_id != KVM_EVMCS_VERSION) &&
+		    (vmx->nested.hv_evmcs->revision_id != VMCS12_REVISION)) {
 			nested_release_evmcs(vcpu);
 			return 0;
 		}
@@ -9390,9 +9420,11 @@ static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu,
 		 * present in struct hv_enlightened_vmcs, ...). Make sure there
 		 * are no leftovers.
 		 */
-		if (from_launch)
-			memset(vmx->nested.cached_vmcs12, 0,
-			       sizeof(*vmx->nested.cached_vmcs12));
+		if (from_launch) {
+			struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
+			memset(vmcs12, 0, sizeof(*vmcs12));
+			vmcs12->hdr.revision_id = VMCS12_REVISION;
+		}
 
 	}
 	return 1;
@@ -15062,7 +15094,7 @@ static struct kvm_x86_ops vmx_x86_ops __ro_after_init = {
 	.has_wbinvd_exit = cpu_has_vmx_wbinvd_exit,
 
 	.read_l1_tsc_offset = vmx_read_l1_tsc_offset,
-	.write_tsc_offset = vmx_write_tsc_offset,
+	.write_l1_tsc_offset = vmx_write_l1_tsc_offset,
 
 	.set_tdp_cr3 = vmx_set_cr3,
 

arch/x86/kvm/x86.c

@@ -1665,8 +1665,7 @@ EXPORT_SYMBOL_GPL(kvm_read_l1_tsc);
 
 static void kvm_vcpu_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
 {
-	kvm_x86_ops->write_tsc_offset(vcpu, offset);
-	vcpu->arch.tsc_offset = offset;
+	vcpu->arch.tsc_offset = kvm_x86_ops->write_l1_tsc_offset(vcpu, offset);
 }
 
 static inline bool kvm_check_tsc_unstable(void)
@@ -1794,7 +1793,8 @@ EXPORT_SYMBOL_GPL(kvm_write_tsc);
 static inline void adjust_tsc_offset_guest(struct kvm_vcpu *vcpu,
 					   s64 adjustment)
 {
-	kvm_vcpu_write_tsc_offset(vcpu, vcpu->arch.tsc_offset + adjustment);
+	u64 tsc_offset = kvm_x86_ops->read_l1_tsc_offset(vcpu);
+	kvm_vcpu_write_tsc_offset(vcpu, tsc_offset + adjustment);
 }
 
 static inline void adjust_tsc_offset_host(struct kvm_vcpu *vcpu, s64 adjustment)
@@ -6918,6 +6918,7 @@ static int kvm_pv_clock_pairing(struct kvm_vcpu *vcpu, gpa_t paddr,
 	clock_pairing.nsec = ts.tv_nsec;
 	clock_pairing.tsc = kvm_read_l1_tsc(vcpu, cycle);
 	clock_pairing.flags = 0;
+	memset(&clock_pairing.pad, 0, sizeof(clock_pairing.pad));
 
 	ret = 0;
 	if (kvm_write_guest(vcpu->kvm, paddr, &clock_pairing,
@@ -7455,7 +7456,8 @@ static void vcpu_scan_ioapic(struct kvm_vcpu *vcpu)
 	else {
 		if (vcpu->arch.apicv_active)
 			kvm_x86_ops->sync_pir_to_irr(vcpu);
-		kvm_ioapic_scan_entry(vcpu, vcpu->arch.ioapic_handled_vectors);
+		if (ioapic_in_kernel(vcpu->kvm))
+			kvm_ioapic_scan_entry(vcpu, vcpu->arch.ioapic_handled_vectors);
 	}
 
 	if (is_guest_mode(vcpu))