Merge branch 'kvm-fixes-for-5.18-rc5' into HEAD

Fixes for (relatively) old bugs, to be merged in both the -rc and next
development trees:

* Fix potential races when walking host page table

* Fix bad user ABI for KVM_EXIT_SYSTEM_EVENT

* Fix shadow page table leak when KVM runs nested

10 files changed, 121 insertions, 32 deletions

diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst
index 85c7abc51af5..4a900cdbc62e 100644
--- a/Documentation/virt/kvm/api.rst
+++ b/Documentation/virt/kvm/api.rst
@@ -5986,16 +5986,16 @@ should put the acknowledged interrupt vector into the 'epr' field.
   #define KVM_SYSTEM_EVENT_RESET          2
   #define KVM_SYSTEM_EVENT_CRASH          3
 			__u32 type;
-			__u64 flags;
+                        __u32 ndata;
+                        __u64 data[16];
 		} system_event;
 
 If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered
 a system-level event using some architecture specific mechanism (hypercall
 or some special instruction). In case of ARM64, this is triggered using
-HVC instruction based PSCI call from the vcpu. The 'type' field describes
-the system-level event type. The 'flags' field describes architecture
-specific flags for the system-level event.
+HVC instruction based PSCI call from the vcpu.
 
+The 'type' field describes the system-level event type.
 Valid values for 'type' are:
 
  - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the
@@ -6010,10 +6010,20 @@ Valid values for 'type' are:
    to ignore the request, or to gather VM memory core dump and/or
    reset/shutdown of the VM.
 
-Valid flags are:
+If KVM_CAP_SYSTEM_EVENT_DATA is present, the 'data' field can contain
+architecture specific information for the system-level event.  Only
+the first `ndata` items (possibly zero) of the data array are valid.
 
- - KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 (arm64 only) -- the guest issued
-   a SYSTEM_RESET2 call according to v1.1 of the PSCI specification.
+ - for arm64, data[0] is set to KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 if
+   the guest issued a SYSTEM_RESET2 call according to v1.1 of the PSCI
+   specification.
+
+ - for RISC-V, data[0] is set to the value of the second argument of the
+   ``sbi_system_reset`` call.
+
+Previous versions of Linux defined a `flags` member in this struct.  The
+field is now aliased to `data[0]`.  Userspace can assume that it is only
+written if ndata is greater than 0.
 
 ::
 
diff --git a/arch/arm64/kvm/psci.c b/arch/arm64/kvm/psci.c
index baac2b405f23..708d80e8e60d 100644
--- a/arch/arm64/kvm/psci.c
+++ b/arch/arm64/kvm/psci.c
@@ -181,7 +181,8 @@ static void kvm_prepare_system_event(struct kvm_vcpu *vcpu, u32 type, u64 flags)
 
 	memset(&vcpu->run->system_event, 0, sizeof(vcpu->run->system_event));
 	vcpu->run->system_event.type = type;
-	vcpu->run->system_event.flags = flags;
+	vcpu->run->system_event.ndata = 1;
+	vcpu->run->system_event.data[0] = flags;
 	vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
 }
 
diff --git a/arch/riscv/kvm/vcpu_sbi.c b/arch/riscv/kvm/vcpu_sbi.c
index a09ecb97b890..d45e7da3f0d3 100644
--- a/arch/riscv/kvm/vcpu_sbi.c
+++ b/arch/riscv/kvm/vcpu_sbi.c
@@ -83,7 +83,7 @@ void kvm_riscv_vcpu_sbi_forward(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 void kvm_riscv_vcpu_sbi_system_reset(struct kvm_vcpu *vcpu,
 				     struct kvm_run *run,
-				     u32 type, u64 flags)
+				     u32 type, u64 reason)
 {
 	unsigned long i;
 	struct kvm_vcpu *tmp;
@@ -94,7 +94,8 @@ void kvm_riscv_vcpu_sbi_system_reset(struct kvm_vcpu *vcpu,
 
 	memset(&run->system_event, 0, sizeof(run->system_event));
 	run->system_event.type = type;
-	run->system_event.flags = flags;
+	run->system_event.ndata = 1;
+	run->system_event.data[0] = reason;
 	run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
 }
 
diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
index e6cae6f22683..a335e7f1f69e 100644
--- a/arch/x86/kvm/mmu.h
+++ b/arch/x86/kvm/mmu.h
@@ -65,6 +65,30 @@ static __always_inline u64 rsvd_bits(int s, int e)
 	return ((2ULL << (e - s)) - 1) << s;
 }
 
+/*
+ * The number of non-reserved physical address bits irrespective of features
+ * that repurpose legal bits, e.g. MKTME.
+ */
+extern u8 __read_mostly shadow_phys_bits;
+
+static inline gfn_t kvm_mmu_max_gfn(void)
+{
+	/*
+	 * Note that this uses the host MAXPHYADDR, not the guest's.
+	 * EPT/NPT cannot support GPAs that would exceed host.MAXPHYADDR;
+	 * assuming KVM is running on bare metal, guest accesses beyond
+	 * host.MAXPHYADDR will hit a #PF(RSVD) and never cause a vmexit
+	 * (either EPT Violation/Misconfig or #NPF), and so KVM will never
+	 * install a SPTE for such addresses.  If KVM is running as a VM
+	 * itself, on the other hand, it might see a MAXPHYADDR that is less
+	 * than hardware's real MAXPHYADDR.  Using the host MAXPHYADDR
+	 * disallows such SPTEs entirely and simplifies the TDP MMU.
+	 */
+	int max_gpa_bits = likely(tdp_enabled) ? shadow_phys_bits : 52;
+
+	return (1ULL << (max_gpa_bits - PAGE_SHIFT)) - 1;
+}
+
 void kvm_mmu_set_mmio_spte_mask(u64 mmio_value, u64 mmio_mask, u64 access_mask);
 void kvm_mmu_set_ept_masks(bool has_ad_bits, bool has_exec_only);
 
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index f9080ee50ffa..64a2a7e2be90 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -2804,8 +2804,12 @@ static int host_pfn_mapping_level(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn,
 				  const struct kvm_memory_slot *slot)
 {
 	unsigned long hva;
-	pte_t *pte;
-	int level;
+	unsigned long flags;
+	int level = PG_LEVEL_4K;
+	pgd_t pgd;
+	p4d_t p4d;
+	pud_t pud;
+	pmd_t pmd;
 
 	if (!PageCompound(pfn_to_page(pfn)) && !kvm_is_zone_device_pfn(pfn))
 		return PG_LEVEL_4K;
@@ -2820,10 +2824,43 @@ static int host_pfn_mapping_level(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn,
 	 */
 	hva = __gfn_to_hva_memslot(slot, gfn);
 
-	pte = lookup_address_in_mm(kvm->mm, hva, &level);
-	if (unlikely(!pte))
-		return PG_LEVEL_4K;
+	/*
+	 * Lookup the mapping level in the current mm.  The information
+	 * may become stale soon, but it is safe to use as long as
+	 * 1) mmu_notifier_retry was checked after taking mmu_lock, and
+	 * 2) mmu_lock is taken now.
+	 *
+	 * We still need to disable IRQs to prevent concurrent tear down
+	 * of page tables.
+	 */
+	local_irq_save(flags);
+
+	pgd = READ_ONCE(*pgd_offset(kvm->mm, hva));
+	if (pgd_none(pgd))
+		goto out;
+
+	p4d = READ_ONCE(*p4d_offset(&pgd, hva));
+	if (p4d_none(p4d) || !p4d_present(p4d))
+		goto out;
 
+	pud = READ_ONCE(*pud_offset(&p4d, hva));
+	if (pud_none(pud) || !pud_present(pud))
+		goto out;
+
+	if (pud_large(pud)) {
+		level = PG_LEVEL_1G;
+		goto out;
+	}
+
+	pmd = READ_ONCE(*pmd_offset(&pud, hva));
+	if (pmd_none(pmd) || !pmd_present(pmd))
+		goto out;
+
+	if (pmd_large(pmd))
+		level = PG_LEVEL_2M;
+
+out:
+	local_irq_restore(flags);
 	return level;
 }
 
@@ -2992,9 +3029,15 @@ static bool handle_abnormal_pfn(struct kvm_vcpu *vcpu, struct kvm_page_fault *fa
 		/*
 		 * If MMIO caching is disabled, emulate immediately without
 		 * touching the shadow page tables as attempting to install an
-		 * MMIO SPTE will just be an expensive nop.
+		 * MMIO SPTE will just be an expensive nop.  Do not cache MMIO
+		 * whose gfn is greater than host.MAXPHYADDR, any guest that
+		 * generates such gfns is running nested and is being tricked
+		 * by L0 userspace (you can observe gfn > L1.MAXPHYADDR if
+		 * and only if L1's MAXPHYADDR is inaccurate with respect to
+		 * the hardware's).
 		 */
-		if (unlikely(!shadow_mmio_value)) {
+		if (unlikely(!shadow_mmio_value) ||
+		    unlikely(fault->gfn > kvm_mmu_max_gfn())) {
 			*ret_val = RET_PF_EMULATE;
 			return true;
 		}
diff --git a/arch/x86/kvm/mmu/spte.h b/arch/x86/kvm/mmu/spte.h
index 73f12615416f..e4abeb5df1b1 100644
--- a/arch/x86/kvm/mmu/spte.h
+++ b/arch/x86/kvm/mmu/spte.h
@@ -201,12 +201,6 @@ static inline bool is_removed_spte(u64 spte)
  */
 extern u64 __read_mostly shadow_nonpresent_or_rsvd_lower_gfn_mask;
 
-/*
- * The number of non-reserved physical address bits irrespective of features
- * that repurpose legal bits, e.g. MKTME.
- */
-extern u8 __read_mostly shadow_phys_bits;
-
 static inline bool is_mmio_spte(u64 spte)
 {
 	return (spte & shadow_mmio_mask) == shadow_mmio_value &&
diff --git a/arch/x86/kvm/mmu/tdp_mmu.c b/arch/x86/kvm/mmu/tdp_mmu.c
index c472769e0300..edc68538819b 100644
--- a/arch/x86/kvm/mmu/tdp_mmu.c
+++ b/arch/x86/kvm/mmu/tdp_mmu.c
@@ -815,14 +815,15 @@ static inline bool __must_check tdp_mmu_iter_cond_resched(struct kvm *kvm,
 	return iter->yielded;
 }
 
-static inline gfn_t tdp_mmu_max_gfn_host(void)
+static inline gfn_t tdp_mmu_max_gfn_exclusive(void)
 {
 	/*
-	 * Bound TDP MMU walks at host.MAXPHYADDR, guest accesses beyond that
-	 * will hit a #PF(RSVD) and never hit an EPT Violation/Misconfig / #NPF,
-	 * and so KVM will never install a SPTE for such addresses.
+	 * Bound TDP MMU walks at host.MAXPHYADDR.  KVM disallows memslots with
+	 * a gpa range that would exceed the max gfn, and KVM does not create
+	 * MMIO SPTEs for "impossible" gfns, instead sending such accesses down
+	 * the slow emulation path every time.
 	 */
-	return 1ULL << (shadow_phys_bits - PAGE_SHIFT);
+	return kvm_mmu_max_gfn() + 1;
 }
 
 static void __tdp_mmu_zap_root(struct kvm *kvm, struct kvm_mmu_page *root,
@@ -830,7 +831,7 @@ static void __tdp_mmu_zap_root(struct kvm *kvm, struct kvm_mmu_page *root,
 {
 	struct tdp_iter iter;
 
-	gfn_t end = tdp_mmu_max_gfn_host();
+	gfn_t end = tdp_mmu_max_gfn_exclusive();
 	gfn_t start = 0;
 
 	for_each_tdp_pte_min_level(iter, root, zap_level, start, end) {
@@ -923,7 +924,7 @@ static bool tdp_mmu_zap_leafs(struct kvm *kvm, struct kvm_mmu_page *root,
 {
 	struct tdp_iter iter;
 
-	end = min(end, tdp_mmu_max_gfn_host());
+	end = min(end, tdp_mmu_max_gfn_exclusive());
 
 	lockdep_assert_held_write(&kvm->mmu_lock);
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a6ab19afc638..4790f0d7d40b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -10020,12 +10020,14 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 		if (kvm_check_request(KVM_REQ_HV_CRASH, vcpu)) {
 			vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
 			vcpu->run->system_event.type = KVM_SYSTEM_EVENT_CRASH;
+			vcpu->run->system_event.ndata = 0;
 			r = 0;
 			goto out;
 		}
 		if (kvm_check_request(KVM_REQ_HV_RESET, vcpu)) {
 			vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
 			vcpu->run->system_event.type = KVM_SYSTEM_EVENT_RESET;
+			vcpu->run->system_event.ndata = 0;
 			r = 0;
 			goto out;
 		}
@@ -12009,8 +12011,12 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
 				   struct kvm_memory_slot *new,
 				   enum kvm_mr_change change)
 {
-	if (change == KVM_MR_CREATE || change == KVM_MR_MOVE)
+	if (change == KVM_MR_CREATE || change == KVM_MR_MOVE) {
+		if ((new->base_gfn + new->npages - 1) > kvm_mmu_max_gfn())
+			return -EINVAL;
+
 		return kvm_alloc_memslot_metadata(kvm, new);
+	}
 
 	if (change == KVM_MR_FLAGS_ONLY)
 		memcpy(&new->arch, &old->arch, sizeof(old->arch));
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index 91a6fe4e02c0..6a184d260c7f 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -445,7 +445,13 @@ struct kvm_run {
 #define KVM_SYSTEM_EVENT_RESET          2
 #define KVM_SYSTEM_EVENT_CRASH          3
 			__u32 type;
-			__u64 flags;
+			__u32 ndata;
+			union {
+#ifndef __KERNEL__
+				__u64 flags;
+#endif
+				__u64 data[16];
+			};
 		} system_event;
 		/* KVM_EXIT_S390_STSI */
 		struct {
@@ -1144,6 +1150,8 @@ struct kvm_ppc_resize_hpt {
 #define KVM_CAP_S390_MEM_OP_EXTENSION 211
 #define KVM_CAP_PMU_CAPABILITY 212
 #define KVM_CAP_DISABLE_QUIRKS2 213
+/* #define KVM_CAP_VM_TSC_CONTROL 214 */
+#define KVM_CAP_SYSTEM_EVENT_DATA 215
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f30bb8c16f26..6d971fb1b08d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -4354,6 +4354,7 @@ static long kvm_vm_ioctl_check_extension_generic(struct kvm *kvm, long arg)
 		return 0;
 #endif
 	case KVM_CAP_BINARY_STATS_FD:
+	case KVM_CAP_SYSTEM_EVENT_DATA:
 		return 1;
 	default:
 		break;