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

Pull more kvm updates from Paolo Bonzini:
 "New x86 features:

   - Guest API and guest kernel support for SEV live migration

   - SEV and SEV-ES intra-host migration

  Bugfixes and cleanups for x86:

   - Fix misuse of gfn-to-pfn cache when recording guest steal time /
     preempted status

   - Fix selftests on APICv machines

   - Fix sparse warnings

   - Fix detection of KVM features in CPUID

   - Cleanups for bogus writes to MSR_KVM_PV_EOI_EN

   - Fixes and cleanups for MSR bitmap handling

   - Cleanups for INVPCID

   - Make x86 KVM_SOFT_MAX_VCPUS consistent with other architectures

  Bugfixes for ARM:

   - Fix finalization of host stage2 mappings

   - Tighten the return value of kvm_vcpu_preferred_target()

   - Make sure the extraction of ESR_ELx.EC is limited to architected
     bits"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (34 commits)
  KVM: SEV: unify cgroup cleanup code for svm_vm_migrate_from
  KVM: x86: move guest_pv_has out of user_access section
  KVM: x86: Drop arbitrary KVM_SOFT_MAX_VCPUS
  KVM: Move INVPCID type check from vmx and svm to the common kvm_handle_invpcid()
  KVM: VMX: Add a helper function to retrieve the GPR index for INVPCID, INVVPID, and INVEPT
  KVM: nVMX: Clean up x2APIC MSR handling for L2
  KVM: VMX: Macrofy the MSR bitmap getters and setters
  KVM: nVMX: Handle dynamic MSR intercept toggling
  KVM: nVMX: Query current VMCS when determining if MSR bitmaps are in use
  KVM: x86: Don't update vcpu->arch.pv_eoi.msr_val when a bogus value was written to MSR_KVM_PV_EOI_EN
  KVM: x86: Rename kvm_lapic_enable_pv_eoi()
  KVM: x86: Make sure KVM_CPUID_FEATURES really are KVM_CPUID_FEATURES
  KVM: x86: Add helper to consolidate core logic of SET_CPUID{2} flows
  kvm: mmu: Use fast PF path for access tracking of huge pages when possible
  KVM: x86/mmu: Properly dereference rcu-protected TDP MMU sptep iterator
  KVM: x86: inhibit APICv when KVM_GUESTDBG_BLOCKIRQ active
  kvm: x86: Convert return type of *is_valid_rdpmc_ecx() to bool
  KVM: x86: Fix recording of guest steal time / preempted status
  selftest: KVM: Add intra host migration tests
  selftest: KVM: Add open sev dev helper
  ...

1 files changed, 112 insertions, 35 deletions

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c1c4e2b05a63..dc7eb5fddfd3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3260,8 +3260,11 @@ static void kvm_vcpu_flush_tlb_guest(struct kvm_vcpu *vcpu)
 
 static void record_steal_time(struct kvm_vcpu *vcpu)
 {
-	struct kvm_host_map map;
-	struct kvm_steal_time *st;
+	struct gfn_to_hva_cache *ghc = &vcpu->arch.st.cache;
+	struct kvm_steal_time __user *st;
+	struct kvm_memslots *slots;
+	u64 steal;
+	u32 version;
 
 	if (kvm_xen_msr_enabled(vcpu->kvm)) {
 		kvm_xen_runstate_set_running(vcpu);
@@ -3271,47 +3274,86 @@ static void record_steal_time(struct kvm_vcpu *vcpu)
 	if (!(vcpu->arch.st.msr_val & KVM_MSR_ENABLED))
 		return;
 
-	/* -EAGAIN is returned in atomic context so we can just return. */
-	if (kvm_map_gfn(vcpu, vcpu->arch.st.msr_val >> PAGE_SHIFT,
-			&map, &vcpu->arch.st.cache, false))
+	if (WARN_ON_ONCE(current->mm != vcpu->kvm->mm))
 		return;
 
-	st = map.hva +
-		offset_in_page(vcpu->arch.st.msr_val & KVM_STEAL_VALID_BITS);
+	slots = kvm_memslots(vcpu->kvm);
+
+	if (unlikely(slots->generation != ghc->generation ||
+		     kvm_is_error_hva(ghc->hva) || !ghc->memslot)) {
+		gfn_t gfn = vcpu->arch.st.msr_val & KVM_STEAL_VALID_BITS;
+
+		/* We rely on the fact that it fits in a single page. */
+		BUILD_BUG_ON((sizeof(*st) - 1) & KVM_STEAL_VALID_BITS);
+
+		if (kvm_gfn_to_hva_cache_init(vcpu->kvm, ghc, gfn, sizeof(*st)) ||
+		    kvm_is_error_hva(ghc->hva) || !ghc->memslot)
+			return;
+	}
 
+	st = (struct kvm_steal_time __user *)ghc->hva;
 	/*
 	 * Doing a TLB flush here, on the guest's behalf, can avoid
 	 * expensive IPIs.
 	 */
 	if (guest_pv_has(vcpu, KVM_FEATURE_PV_TLB_FLUSH)) {
-		u8 st_preempted = xchg(&st->preempted, 0);
+		u8 st_preempted = 0;
+		int err = -EFAULT;
+
+		if (!user_access_begin(st, sizeof(*st)))
+			return;
+
+		asm volatile("1: xchgb %0, %2\n"
+			     "xor %1, %1\n"
+			     "2:\n"
+			     _ASM_EXTABLE_UA(1b, 2b)
+			     : "+r" (st_preempted),
+			       "+&r" (err)
+			     : "m" (st->preempted));
+		if (err)
+			goto out;
+
+		user_access_end();
+
+		vcpu->arch.st.preempted = 0;
 
 		trace_kvm_pv_tlb_flush(vcpu->vcpu_id,
 				       st_preempted & KVM_VCPU_FLUSH_TLB);
 		if (st_preempted & KVM_VCPU_FLUSH_TLB)
 			kvm_vcpu_flush_tlb_guest(vcpu);
+
+		if (!user_access_begin(st, sizeof(*st)))
+			goto dirty;
 	} else {
-		st->preempted = 0;
-	}
+		if (!user_access_begin(st, sizeof(*st)))
+			return;
 
-	vcpu->arch.st.preempted = 0;
+		unsafe_put_user(0, &st->preempted, out);
+		vcpu->arch.st.preempted = 0;
+	}
 
-	if (st->version & 1)
-		st->version += 1;  /* first time write, random junk */
+	unsafe_get_user(version, &st->version, out);
+	if (version & 1)
+		version += 1;  /* first time write, random junk */
 
-	st->version += 1;
+	version += 1;
+	unsafe_put_user(version, &st->version, out);
 
 	smp_wmb();
 
-	st->steal += current->sched_info.run_delay -
+	unsafe_get_user(steal, &st->steal, out);
+	steal += current->sched_info.run_delay -
 		vcpu->arch.st.last_steal;
 	vcpu->arch.st.last_steal = current->sched_info.run_delay;
+	unsafe_put_user(steal, &st->steal, out);
 
-	smp_wmb();
-
-	st->version += 1;
+	version += 1;
+	unsafe_put_user(version, &st->version, out);
 
-	kvm_unmap_gfn(vcpu, &map, &vcpu->arch.st.cache, true, false);
+ out:
+	user_access_end();
+ dirty:
+	mark_page_dirty_in_slot(vcpu->kvm, ghc->memslot, gpa_to_gfn(ghc->gpa));
 }
 
 int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
@@ -3517,7 +3559,7 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 		if (!guest_pv_has(vcpu, KVM_FEATURE_PV_EOI))
 			return 1;
 
-		if (kvm_lapic_enable_pv_eoi(vcpu, data, sizeof(u8)))
+		if (kvm_lapic_set_pv_eoi(vcpu, data, sizeof(u8)))
 			return 1;
 		break;
 
@@ -4137,7 +4179,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 		r = !static_call(kvm_x86_cpu_has_accelerated_tpr)();
 		break;
 	case KVM_CAP_NR_VCPUS:
-		r = KVM_SOFT_MAX_VCPUS;
+		r = num_online_cpus();
 		break;
 	case KVM_CAP_MAX_VCPUS:
 		r = KVM_MAX_VCPUS;
@@ -4351,8 +4393,10 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 
 static void kvm_steal_time_set_preempted(struct kvm_vcpu *vcpu)
 {
-	struct kvm_host_map map;
-	struct kvm_steal_time *st;
+	struct gfn_to_hva_cache *ghc = &vcpu->arch.st.cache;
+	struct kvm_steal_time __user *st;
+	struct kvm_memslots *slots;
+	static const u8 preempted = KVM_VCPU_PREEMPTED;
 
 	if (!(vcpu->arch.st.msr_val & KVM_MSR_ENABLED))
 		return;
@@ -4360,16 +4404,23 @@ static void kvm_steal_time_set_preempted(struct kvm_vcpu *vcpu)
 	if (vcpu->arch.st.preempted)
 		return;
 
-	if (kvm_map_gfn(vcpu, vcpu->arch.st.msr_val >> PAGE_SHIFT, &map,
-			&vcpu->arch.st.cache, true))
+	/* This happens on process exit */
+	if (unlikely(current->mm != vcpu->kvm->mm))
 		return;
 
-	st = map.hva +
-		offset_in_page(vcpu->arch.st.msr_val & KVM_STEAL_VALID_BITS);
+	slots = kvm_memslots(vcpu->kvm);
+
+	if (unlikely(slots->generation != ghc->generation ||
+		     kvm_is_error_hva(ghc->hva) || !ghc->memslot))
+		return;
 
-	st->preempted = vcpu->arch.st.preempted = KVM_VCPU_PREEMPTED;
+	st = (struct kvm_steal_time __user *)ghc->hva;
+	BUILD_BUG_ON(sizeof(st->preempted) != sizeof(preempted));
 
-	kvm_unmap_gfn(vcpu, &map, &vcpu->arch.st.cache, true, true);
+	if (!copy_to_user_nofault(&st->preempted, &preempted, sizeof(preempted)))
+		vcpu->arch.st.preempted = KVM_VCPU_PREEMPTED;
+
+	mark_page_dirty_in_slot(vcpu->kvm, ghc->memslot, gpa_to_gfn(ghc->gpa));
 }
 
 void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
@@ -5728,6 +5779,12 @@ split_irqchip_unlock:
 		if (kvm_x86_ops.vm_copy_enc_context_from)
 			r = kvm_x86_ops.vm_copy_enc_context_from(kvm, cap->args[0]);
 		return r;
+	case KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM:
+		r = -EINVAL;
+		if (kvm_x86_ops.vm_move_enc_context_from)
+			r = kvm_x86_ops.vm_move_enc_context_from(
+				kvm, cap->args[0]);
+		return r;
 	case KVM_CAP_EXIT_HYPERCALL:
 		if (cap->args[0] & ~KVM_EXIT_HYPERCALL_VALID_MASK) {
 			r = -EINVAL;
@@ -7328,7 +7385,9 @@ static void emulator_set_smbase(struct x86_emulate_ctxt *ctxt, u64 smbase)
 static int emulator_check_pmc(struct x86_emulate_ctxt *ctxt,
 			      u32 pmc)
 {
-	return kvm_pmu_is_valid_rdpmc_ecx(emul_to_vcpu(ctxt), pmc);
+	if (kvm_pmu_is_valid_rdpmc_ecx(emul_to_vcpu(ctxt), pmc))
+		return 0;
+	return -EINVAL;
 }
 
 static int emulator_read_pmc(struct x86_emulate_ctxt *ctxt,
@@ -9552,7 +9611,7 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 	}
 
 	if (kvm_request_pending(vcpu)) {
-		if (kvm_check_request(KVM_REQ_VM_BUGGED, vcpu)) {
+		if (kvm_check_request(KVM_REQ_VM_DEAD, vcpu)) {
 			r = -EIO;
 			goto out;
 		}
@@ -10564,6 +10623,24 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	return ret;
 }
 
+static void kvm_arch_vcpu_guestdbg_update_apicv_inhibit(struct kvm *kvm)
+{
+	bool inhibit = false;
+	struct kvm_vcpu *vcpu;
+	int i;
+
+	down_write(&kvm->arch.apicv_update_lock);
+
+	kvm_for_each_vcpu(i, vcpu, kvm) {
+		if (vcpu->guest_debug & KVM_GUESTDBG_BLOCKIRQ) {
+			inhibit = true;
+			break;
+		}
+	}
+	__kvm_request_apicv_update(kvm, !inhibit, APICV_INHIBIT_REASON_BLOCKIRQ);
+	up_write(&kvm->arch.apicv_update_lock);
+}
+
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
@@ -10616,6 +10693,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 
 	static_call(kvm_x86_update_exception_bitmap)(vcpu);
 
+	kvm_arch_vcpu_guestdbg_update_apicv_inhibit(vcpu->kvm);
+
 	r = 0;
 
 out:
@@ -10859,11 +10938,8 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
-	struct gfn_to_pfn_cache *cache = &vcpu->arch.st.cache;
 	int idx;
 
-	kvm_release_pfn(cache->pfn, cache->dirty, cache);
-
 	kvmclock_reset(vcpu);
 
 	static_call(kvm_x86_vcpu_free)(vcpu);
@@ -12275,7 +12351,8 @@ int kvm_handle_invpcid(struct kvm_vcpu *vcpu, unsigned long type, gva_t gva)
 		return kvm_skip_emulated_instruction(vcpu);
 
 	default:
-		BUG(); /* We have already checked above that type <= 3 */
+		kvm_inject_gp(vcpu, 0);
+		return 1;
 	}
 }
 EXPORT_SYMBOL_GPL(kvm_handle_invpcid);