KVM x86 misc changes for 6.11

- Add a global struct to consolidate tracking of host values, e.g. EFER, and
    move "shadow_phys_bits" into the structure as "maxphyaddr".
 
  - Add KVM_CAP_X86_APIC_BUS_CYCLES_NS to allow configuring the effective APIC
    bus frequency, because TDX.
 
  - Print the name of the APICv/AVIC inhibits in the relevant tracepoint.
 
  - Clean up KVM's handling of vendor specific emulation to consistently act on
    "compatible with Intel/AMD", versus checking for a specific vendor.
 
  - Misc cleanups
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEKTobbabEP7vbhhN9OlYIJqCjN/0FAmaRub0ACgkQOlYIJqCj
 N/2LMxAArGzhcWZ6Qdo2aMRaMIPtSBJHmbEgEuHvHMumgsTZQzDcn9cxDi/hNSrc
 l8ODOwAM2qNcq95YfwjU7F0ae3E+HRzGvKcBnmZWuQeCDp2HhVEoCphFu1sHst+t
 XEJTL02b6OgyJUEU3h40mYk12eiq2S4FCnFYXPCqijwwuL6Y5KQvvTqek3c2/SDn
 c+VneutYGax/S0GiiCkYh4wrwWh9g7qm0IX70ycBwJbW5qBFKgyglvHxvL8JLJC9
 Nkkw/p2657wcOdraH+fOBuRy2dMwE5fv++1tOjWwB5WAAhSOJPZh0BGYvgA2yfN7
 OE+k7APKUQd9Xxtud8H3LrTPoyMA4hz2sdDFyqrrWK9yjpBY7zXNyN50Fxi7VVsm
 T8nTIiKAGyRbjotY+m7krXQPXjfZYhVqrJ/jtxESOZLZ93q2gSWU2p/ZXpUPVHnH
 +YOBAI1owP3wepaYlrthtI4LQx9lF422dnmeSflztfKFGabRbQZxg3uHMCCxIaGc
 lJ6CD546+D45f/uBXRDMqk//qFTqXhKUbDk9sutmU/C2oWufMwW0R8kOyItGPyvk
 9PP1vd8vSsIHj+tpwg+i04jBqYDaAcPBOcTZaHm9SYYP+1e11Uu5Vjep37JL1bkA
 xJWxnDZOCGcfKQi2jkh51HJ/dOAHXY1GQKMfyAoPQOSonYHvGVY=
 =Cf2R
 -----END PGP SIGNATURE-----

Merge tag 'kvm-x86-misc-6.11' of https://github.com/kvm-x86/linux into HEAD

KVM x86 misc changes for 6.11

 - Add a global struct to consolidate tracking of host values, e.g. EFER, and
   move "shadow_phys_bits" into the structure as "maxphyaddr".

 - Add KVM_CAP_X86_APIC_BUS_CYCLES_NS to allow configuring the effective APIC
   bus frequency, because TDX.

 - Print the name of the APICv/AVIC inhibits in the relevant tracepoint.

 - Clean up KVM's handling of vendor specific emulation to consistently act on
   "compatible with Intel/AMD", versus checking for a specific vendor.

 - Misc cleanups

Documentation/virt/kvm/api.rst

@@ -6483,9 +6483,12 @@ More architecture-specific flags detailing state of the VCPU that may
 affect the device's behavior. Current defined flags::
 
   /* x86, set if the VCPU is in system management mode */
-  #define KVM_RUN_X86_SMM     (1 << 0)
+  #define KVM_RUN_X86_SMM          (1 << 0)
   /* x86, set if bus lock detected in VM */
-  #define KVM_RUN_BUS_LOCK    (1 << 1)
+  #define KVM_RUN_X86_BUS_LOCK     (1 << 1)
+  /* x86, set if the VCPU is executing a nested (L2) guest */
+  #define KVM_RUN_X86_GUEST_MODE   (1 << 2)
+
   /* arm64, set for KVM_EXIT_DEBUG */
   #define KVM_DEBUG_ARCH_HSR_HIGH_VALID  (1 << 0)
 
@@ -7831,29 +7834,31 @@ Valid bits in args[0] are::
   #define KVM_BUS_LOCK_DETECTION_OFF      (1 << 0)
   #define KVM_BUS_LOCK_DETECTION_EXIT     (1 << 1)
 
-Enabling this capability on a VM provides userspace with a way to select
+Enabling this capability on a VM provides userspace with a way to select a
-a policy to handle the bus locks detected in guest. Userspace can obtain
+policy to handle the bus locks detected in guest. Userspace can obtain the
-the supported modes from the result of KVM_CHECK_EXTENSION and define it
+supported modes from the result of KVM_CHECK_EXTENSION and define it through
-through the KVM_ENABLE_CAP.
+the KVM_ENABLE_CAP. The supported modes are mutually-exclusive.
 
-KVM_BUS_LOCK_DETECTION_OFF and KVM_BUS_LOCK_DETECTION_EXIT are supported
+This capability allows userspace to force VM exits on bus locks detected in the
-currently and mutually exclusive with each other. More bits can be added in
+guest, irrespective whether or not the host has enabled split-lock detection
-the future.
+(which triggers an #AC exception that KVM intercepts). This capability is
+intended to mitigate attacks where a malicious/buggy guest can exploit bus
+locks to degrade the performance of the whole system.
 
-With KVM_BUS_LOCK_DETECTION_OFF set, bus locks in guest will not cause vm exits
+If KVM_BUS_LOCK_DETECTION_OFF is set, KVM doesn't force guest bus locks to VM
-so that no additional actions are needed. This is the default mode.
+exit, although the host kernel's split-lock #AC detection still applies, if
+enabled.
 
-With KVM_BUS_LOCK_DETECTION_EXIT set, vm exits happen when bus lock detected
+If KVM_BUS_LOCK_DETECTION_EXIT is set, KVM enables a CPU feature that ensures
-in VM. KVM just exits to userspace when handling them. Userspace can enforce
+bus locks in the guest trigger a VM exit, and KVM exits to userspace for all
-its own throttling or other policy based mitigations.
+such VM exits, e.g. to allow userspace to throttle the offending guest and/or
+apply some other policy-based mitigation. When exiting to userspace, KVM sets
+KVM_RUN_X86_BUS_LOCK in vcpu-run->flags, and conditionally sets the exit_reason
+to KVM_EXIT_X86_BUS_LOCK.
 
-This capability is aimed to address the thread that VM can exploit bus locks to
+Note! Detected bus locks may be coincident with other exits to userspace, i.e.
-degree the performance of the whole system. Once the userspace enable this
+KVM_RUN_X86_BUS_LOCK should be checked regardless of the primary exit reason if
-capability and select the KVM_BUS_LOCK_DETECTION_EXIT mode, KVM will set the
+userspace wants to take action on all detected bus locks.
-KVM_RUN_BUS_LOCK flag in vcpu-run->flags field and exit to userspace. Concerning
-the bus lock vm exit can be preempted by a higher priority VM exit, the exit
-notifications to userspace can be KVM_EXIT_BUS_LOCK or other reasons.
-KVM_RUN_BUS_LOCK flag is used to distinguish between them.
 
 7.23 KVM_CAP_PPC_DAWR1
 ----------------------
@@ -8137,6 +8142,37 @@ error/annotated fault.
 
 See KVM_EXIT_MEMORY_FAULT for more information.
 
+7.35 KVM_CAP_X86_APIC_BUS_CYCLES_NS
+-----------------------------------
+
+:Architectures: x86
+:Target: VM
+:Parameters: args[0] is the desired APIC bus clock rate, in nanoseconds
+:Returns: 0 on success, -EINVAL if args[0] contains an invalid value for the
+          frequency or if any vCPUs have been created, -ENXIO if a virtual
+          local APIC has not been created using KVM_CREATE_IRQCHIP.
+
+This capability sets the VM's APIC bus clock frequency, used by KVM's in-kernel
+virtual APIC when emulating APIC timers.  KVM's default value can be retrieved
+by KVM_CHECK_EXTENSION.
+
+Note: Userspace is responsible for correctly configuring CPUID 0x15, a.k.a. the
+core crystal clock frequency, if a non-zero CPUID 0x15 is exposed to the guest.
+
+7.36 KVM_CAP_X86_GUEST_MODE
+------------------------------
+
+:Architectures: x86
+:Returns: Informational only, -EINVAL on direct KVM_ENABLE_CAP.
+
+The presence of this capability indicates that KVM_RUN will update the
+KVM_RUN_X86_GUEST_MODE bit in kvm_run.flags to indicate whether the
+vCPU was executing nested guest code when it exited.
+
+KVM exits with the register state of either the L1 or L2 guest
+depending on which executed at the time of an exit. Userspace must
+take care to differentiate between these cases.
+
 8. Other capabilities.
 ======================
 

arch/x86/include/asm/kvm_host.h

@@ -1208,7 +1208,7 @@ enum kvm_apicv_inhibit {
 	 * APIC acceleration is disabled by a module parameter
 	 * and/or not supported in hardware.
 	 */
-	APICV_INHIBIT_REASON_DISABLE,
+	APICV_INHIBIT_REASON_DISABLED,
 
 	/*
 	 * APIC acceleration is inhibited because AutoEOI feature is
@@ -1278,8 +1278,27 @@ enum kvm_apicv_inhibit {
 	 * mapping between logical ID and vCPU.
 	 */
 	APICV_INHIBIT_REASON_LOGICAL_ID_ALIASED,
+
+	NR_APICV_INHIBIT_REASONS,
 };
 
+#define __APICV_INHIBIT_REASON(reason)			\
+	{ BIT(APICV_INHIBIT_REASON_##reason), #reason }
+
+#define APICV_INHIBIT_REASONS				\
+	__APICV_INHIBIT_REASON(DISABLED),		\
+	__APICV_INHIBIT_REASON(HYPERV),			\
+	__APICV_INHIBIT_REASON(ABSENT),			\
+	__APICV_INHIBIT_REASON(BLOCKIRQ),		\
+	__APICV_INHIBIT_REASON(PHYSICAL_ID_ALIASED),	\
+	__APICV_INHIBIT_REASON(APIC_ID_MODIFIED),	\
+	__APICV_INHIBIT_REASON(APIC_BASE_MODIFIED),	\
+	__APICV_INHIBIT_REASON(NESTED),			\
+	__APICV_INHIBIT_REASON(IRQWIN),			\
+	__APICV_INHIBIT_REASON(PIT_REINJ),		\
+	__APICV_INHIBIT_REASON(SEV),			\
+	__APICV_INHIBIT_REASON(LOGICAL_ID_ALIASED)
+
 struct kvm_arch {
 	unsigned long n_used_mmu_pages;
 	unsigned long n_requested_mmu_pages;
@@ -1365,6 +1384,7 @@ struct kvm_arch {
 
 	u32 default_tsc_khz;
 	bool user_set_tsc;
+	u64 apic_bus_cycle_ns;
 
 	seqcount_raw_spinlock_t pvclock_sc;
 	bool use_master_clock;
@@ -1709,7 +1729,6 @@ struct kvm_x86_ops {
 	void (*enable_nmi_window)(struct kvm_vcpu *vcpu);
 	void (*enable_irq_window)(struct kvm_vcpu *vcpu);
 	void (*update_cr8_intercept)(struct kvm_vcpu *vcpu, int tpr, int irr);
-	bool (*check_apicv_inhibit_reasons)(enum kvm_apicv_inhibit reason);
 	const unsigned long required_apicv_inhibits;
 	bool allow_apicv_in_x2apic_without_x2apic_virtualization;
 	void (*refresh_apicv_exec_ctrl)(struct kvm_vcpu *vcpu);
@@ -1855,7 +1874,6 @@ struct kvm_arch_async_pf {
 };
 
 extern u32 __read_mostly kvm_nr_uret_msrs;
-extern u64 __read_mostly host_efer;
 extern bool __read_mostly allow_smaller_maxphyaddr;
 extern bool __read_mostly enable_apicv;
 extern struct kvm_x86_ops kvm_x86_ops;

arch/x86/include/uapi/asm/kvm.h

@@ -106,6 +106,7 @@ struct kvm_ioapic_state {
 
 #define KVM_RUN_X86_SMM		 (1 << 0)
 #define KVM_RUN_X86_BUS_LOCK     (1 << 1)
+#define KVM_RUN_X86_GUEST_MODE   (1 << 2)
 
 /* for KVM_GET_REGS and KVM_SET_REGS */
 struct kvm_regs {

arch/x86/kvm/cpuid.c

@@ -335,6 +335,18 @@ static bool kvm_cpuid_has_hyperv(struct kvm_cpuid_entry2 *entries, int nent)
 #endif
 }
 
+static bool guest_cpuid_is_amd_or_hygon(struct kvm_vcpu *vcpu)
+{
+	struct kvm_cpuid_entry2 *entry;
+
+	entry = kvm_find_cpuid_entry(vcpu, 0);
+	if (!entry)
+		return false;
+
+	return is_guest_vendor_amd(entry->ebx, entry->ecx, entry->edx) ||
+	       is_guest_vendor_hygon(entry->ebx, entry->ecx, entry->edx);
+}
+
 static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
 {
 	struct kvm_lapic *apic = vcpu->arch.apic;

arch/x86/kvm/cpuid.h

@@ -102,24 +102,6 @@ static __always_inline void guest_cpuid_clear(struct kvm_vcpu *vcpu,
 		*reg &= ~__feature_bit(x86_feature);
 }
 
-static inline bool guest_cpuid_is_amd_or_hygon(struct kvm_vcpu *vcpu)
-{
-	struct kvm_cpuid_entry2 *best;
-
-	best = kvm_find_cpuid_entry(vcpu, 0);
-	return best &&
-	       (is_guest_vendor_amd(best->ebx, best->ecx, best->edx) ||
-		is_guest_vendor_hygon(best->ebx, best->ecx, best->edx));
-}
-
-static inline bool guest_cpuid_is_intel(struct kvm_vcpu *vcpu)
-{
-	struct kvm_cpuid_entry2 *best;
-
-	best = kvm_find_cpuid_entry(vcpu, 0);
-	return best && is_guest_vendor_intel(best->ebx, best->ecx, best->edx);
-}
-
 static inline bool guest_cpuid_is_amd_compatible(struct kvm_vcpu *vcpu)
 {
 	return vcpu->arch.is_amd_compatible;

arch/x86/kvm/emulate.c

@@ -2354,50 +2354,6 @@ setup_syscalls_segments(struct desc_struct *cs, struct desc_struct *ss)
 	ss->avl = 0;
 }
 
-static bool vendor_intel(struct x86_emulate_ctxt *ctxt)
-{
-	u32 eax, ebx, ecx, edx;
-
-	eax = ecx = 0;
-	ctxt->ops->get_cpuid(ctxt, &eax, &ebx, &ecx, &edx, true);
-	return is_guest_vendor_intel(ebx, ecx, edx);
-}
-
-static bool em_syscall_is_enabled(struct x86_emulate_ctxt *ctxt)
-{
-	const struct x86_emulate_ops *ops = ctxt->ops;
-	u32 eax, ebx, ecx, edx;
-
-	/*
-	 * syscall should always be enabled in longmode - so only become
-	 * vendor specific (cpuid) if other modes are active...
-	 */
-	if (ctxt->mode == X86EMUL_MODE_PROT64)
-		return true;
-
-	eax = 0x00000000;
-	ecx = 0x00000000;
-	ops->get_cpuid(ctxt, &eax, &ebx, &ecx, &edx, true);
-	/*
-	 * remark: Intel CPUs only support "syscall" in 64bit longmode. Also a
-	 * 64bit guest with a 32bit compat-app running will #UD !! While this
-	 * behaviour can be fixed (by emulating) into AMD response - CPUs of
-	 * AMD can't behave like Intel.
-	 */
-	if (is_guest_vendor_intel(ebx, ecx, edx))
-		return false;
-
-	if (is_guest_vendor_amd(ebx, ecx, edx) ||
-	    is_guest_vendor_hygon(ebx, ecx, edx))
-		return true;
-
-	/*
-	 * default: (not Intel, not AMD, not Hygon), apply Intel's
-	 * stricter rules...
-	 */
-	return false;
-}
-
 static int em_syscall(struct x86_emulate_ctxt *ctxt)
 {
 	const struct x86_emulate_ops *ops = ctxt->ops;
@@ -2411,7 +2367,15 @@ static int em_syscall(struct x86_emulate_ctxt *ctxt)
 	    ctxt->mode == X86EMUL_MODE_VM86)
 		return emulate_ud(ctxt);
 
-	if (!(em_syscall_is_enabled(ctxt)))
+	/*
+	 * Intel compatible CPUs only support SYSCALL in 64-bit mode, whereas
+	 * AMD allows SYSCALL in any flavor of protected mode.  Note, it's
+	 * infeasible to emulate Intel behavior when running on AMD hardware,
+	 * as SYSCALL won't fault in the "wrong" mode, i.e. there is no #UD
+	 * for KVM to trap-and-emulate, unlike emulating AMD on Intel.
+	 */
+	if (ctxt->mode != X86EMUL_MODE_PROT64 &&
+	    ctxt->ops->guest_cpuid_is_intel_compatible(ctxt))
 		return emulate_ud(ctxt);
 
 	ops->get_msr(ctxt, MSR_EFER, &efer);
@@ -2471,11 +2435,11 @@ static int em_sysenter(struct x86_emulate_ctxt *ctxt)
 		return emulate_gp(ctxt, 0);
 
 	/*
-	 * Not recognized on AMD in compat mode (but is recognized in legacy
+	 * Intel's architecture allows SYSENTER in compatibility mode, but AMD
-	 * mode).
+	 * does not.  Note, AMD does allow SYSENTER in legacy protected mode.
 	 */
-	if ((ctxt->mode != X86EMUL_MODE_PROT64) && (efer & EFER_LMA)
+	if ((ctxt->mode != X86EMUL_MODE_PROT64) && (efer & EFER_LMA) &&
-	    && !vendor_intel(ctxt))
+	    !ctxt->ops->guest_cpuid_is_intel_compatible(ctxt))
 		return emulate_ud(ctxt);
 
 	/* sysenter/sysexit have not been tested in 64bit mode. */
@@ -2647,7 +2611,14 @@ static void string_registers_quirk(struct x86_emulate_ctxt *ctxt)
 	 * manner when ECX is zero due to REP-string optimizations.
 	 */
 #ifdef CONFIG_X86_64
-	if (ctxt->ad_bytes != 4 || !vendor_intel(ctxt))
+	u32 eax, ebx, ecx, edx;
+
+	if (ctxt->ad_bytes != 4)
+		return;
+
+	eax = ecx = 0;
+	ctxt->ops->get_cpuid(ctxt, &eax, &ebx, &ecx, &edx, true);
+	if (!is_guest_vendor_intel(ebx, ecx, edx))
 		return;
 
 	*reg_write(ctxt, VCPU_REGS_RCX) = 0;

arch/x86/kvm/hyperv.c

@@ -1737,7 +1737,8 @@ static int kvm_hv_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata,
 		data = (u64)vcpu->arch.virtual_tsc_khz * 1000;
 		break;
 	case HV_X64_MSR_APIC_FREQUENCY:
-		data = APIC_BUS_FREQUENCY;
+		data = div64_u64(1000000000ULL,
+				 vcpu->kvm->arch.apic_bus_cycle_ns);
 		break;
 	default:
 		kvm_pr_unimpl_rdmsr(vcpu, msr);

arch/x86/kvm/kvm_emulate.h

@@ -223,6 +223,7 @@ struct x86_emulate_ops {
 	bool (*guest_has_movbe)(struct x86_emulate_ctxt *ctxt);
 	bool (*guest_has_fxsr)(struct x86_emulate_ctxt *ctxt);
 	bool (*guest_has_rdpid)(struct x86_emulate_ctxt *ctxt);
+	bool (*guest_cpuid_is_intel_compatible)(struct x86_emulate_ctxt *ctxt);
 
 	void (*set_nmi_mask)(struct x86_emulate_ctxt *ctxt, bool masked);
 

arch/x86/kvm/lapic.c

@@ -1557,7 +1557,8 @@ static u32 apic_get_tmcct(struct kvm_lapic *apic)
 		remaining = 0;
 
 	ns = mod_64(ktime_to_ns(remaining), apic->lapic_timer.period);
-	return div64_u64(ns, (APIC_BUS_CYCLE_NS * apic->divide_count));
+	return div64_u64(ns, (apic->vcpu->kvm->arch.apic_bus_cycle_ns *
+			      apic->divide_count));
 }
 
 static void __report_tpr_access(struct kvm_lapic *apic, bool write)
@@ -1973,7 +1974,8 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
 
 static inline u64 tmict_to_ns(struct kvm_lapic *apic, u32 tmict)
 {
-	return (u64)tmict * APIC_BUS_CYCLE_NS * (u64)apic->divide_count;
+	return (u64)tmict * apic->vcpu->kvm->arch.apic_bus_cycle_ns *
+		(u64)apic->divide_count;
 }
 
 static void update_target_expiration(struct kvm_lapic *apic, uint32_t old_divisor)

arch/x86/kvm/lapic.h

@@ -16,8 +16,7 @@
 #define APIC_DEST_NOSHORT		0x0
 #define APIC_DEST_MASK			0x800
 
-#define APIC_BUS_CYCLE_NS       1
+#define APIC_BUS_CYCLE_NS_DEFAULT	1
-#define APIC_BUS_FREQUENCY      (1000000000ULL / APIC_BUS_CYCLE_NS)
 
 #define APIC_BROADCAST			0xFF
 #define X2APIC_BROADCAST		0xFFFFFFFFul

arch/x86/kvm/mmu.h

@@ -57,12 +57,6 @@ 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)
 {
 	/*
@@ -76,30 +70,11 @@ static inline gfn_t kvm_mmu_max_gfn(void)
 	 * 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;
+	int max_gpa_bits = likely(tdp_enabled) ? kvm_host.maxphyaddr : 52;
 
 	return (1ULL << (max_gpa_bits - PAGE_SHIFT)) - 1;
 }
 
-static inline u8 kvm_get_shadow_phys_bits(void)
-{
-	/*
-	 * boot_cpu_data.x86_phys_bits is reduced when MKTME or SME are detected
-	 * in CPU detection code, but the processor treats those reduced bits as
-	 * 'keyID' thus they are not reserved bits. Therefore KVM needs to look at
-	 * the physical address bits reported by CPUID.
-	 */
-	if (likely(boot_cpu_data.extended_cpuid_level >= 0x80000008))
-		return cpuid_eax(0x80000008) & 0xff;
-
-	/*
-	 * Quite weird to have VMX or SVM but not MAXPHYADDR; probably a VM with
-	 * custom CPUID.  Proceed with whatever the kernel found since these features
-	 * aren't virtualizable (SME/SEV also require CPUIDs higher than 0x80000008).
-	 */
-	return boot_cpu_data.x86_phys_bits;
-}
-
 u8 kvm_mmu_get_max_tdp_level(void);
 
 void kvm_mmu_set_mmio_spte_mask(u64 mmio_value, u64 mmio_mask, u64 access_mask);

arch/x86/kvm/mmu/mmu.c

@@ -5109,7 +5109,7 @@ static void reset_rsvds_bits_mask_ept(struct kvm_vcpu *vcpu,
 
 static inline u64 reserved_hpa_bits(void)
 {
-	return rsvd_bits(shadow_phys_bits, 63);
+	return rsvd_bits(kvm_host.maxphyaddr, 63);
 }
 
 /*

arch/x86/kvm/mmu/spte.c

@@ -43,7 +43,25 @@ u64 __read_mostly shadow_acc_track_mask;
 u64 __read_mostly shadow_nonpresent_or_rsvd_mask;
 u64 __read_mostly shadow_nonpresent_or_rsvd_lower_gfn_mask;
 
-u8 __read_mostly shadow_phys_bits;
+static u8 __init kvm_get_host_maxphyaddr(void)
+{
+	/*
+	 * boot_cpu_data.x86_phys_bits is reduced when MKTME or SME are detected
+	 * in CPU detection code, but the processor treats those reduced bits as
+	 * 'keyID' thus they are not reserved bits. Therefore KVM needs to look at
+	 * the physical address bits reported by CPUID, i.e. the raw MAXPHYADDR,
+	 * when reasoning about CPU behavior with respect to MAXPHYADDR.
+	 */
+	if (likely(boot_cpu_data.extended_cpuid_level >= 0x80000008))
+		return cpuid_eax(0x80000008) & 0xff;
+
+	/*
+	 * Quite weird to have VMX or SVM but not MAXPHYADDR; probably a VM with
+	 * custom CPUID.  Proceed with whatever the kernel found since these features
+	 * aren't virtualizable (SME/SEV also require CPUIDs higher than 0x80000008).
+	 */
+	return boot_cpu_data.x86_phys_bits;
+}
 
 void __init kvm_mmu_spte_module_init(void)
 {
@@ -55,6 +73,8 @@ void __init kvm_mmu_spte_module_init(void)
 	 * will change when the vendor module is (re)loaded.
 	 */
 	allow_mmio_caching = enable_mmio_caching;
+
+	kvm_host.maxphyaddr = kvm_get_host_maxphyaddr();
 }
 
 static u64 generation_mmio_spte_mask(u64 gen)
@@ -441,8 +461,6 @@ void kvm_mmu_reset_all_pte_masks(void)
 	u8 low_phys_bits;
 	u64 mask;
 
-	shadow_phys_bits = kvm_get_shadow_phys_bits();
-
 	/*
 	 * If the CPU has 46 or less physical address bits, then set an
 	 * appropriate mask to guard against L1TF attacks. Otherwise, it is
@@ -494,7 +512,7 @@ void kvm_mmu_reset_all_pte_masks(void)
 	 * 52-bit physical addresses then there are no reserved PA bits in the
 	 * PTEs and so the reserved PA approach must be disabled.
 	 */
-	if (shadow_phys_bits < 52)
+	if (kvm_host.maxphyaddr < 52)
 		mask = BIT_ULL(51) | PT_PRESENT_MASK;
 	else
 		mask = 0;

arch/x86/kvm/pmu.c

@@ -194,7 +194,7 @@ static int pmc_reprogram_counter(struct kvm_pmc *pmc, u32 type, u64 config,
 	attr.sample_period = get_sample_period(pmc, pmc->counter);
 
 	if ((attr.config & HSW_IN_TX_CHECKPOINTED) &&
-	    guest_cpuid_is_intel(pmc->vcpu)) {
+	    (boot_cpu_has(X86_FEATURE_RTM) || boot_cpu_has(X86_FEATURE_HLE))) {
 		/*
 		 * HSW_IN_TX_CHECKPOINTED is not supported with nonzero
 		 * period. Just clear the sample period so at least

arch/x86/kvm/svm/sev.c

@@ -4401,9 +4401,9 @@ void sev_es_prepare_switch_to_guest(struct vcpu_svm *svm, struct sev_es_save_are
 	 * isn't saved by VMRUN, that isn't already saved by VMSAVE (performed
 	 * by common SVM code).
 	 */
-	hostsa->xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
+	hostsa->xcr0 = kvm_host.xcr0;
 	hostsa->pkru = read_pkru();
-	hostsa->xss = host_xss;
+	hostsa->xss = kvm_host.xss;
 
 	/*
 	 * If DebugSwap is enabled, debug registers are loaded but NOT saved by

arch/x86/kvm/svm/svm.c

@@ -53,6 +53,7 @@
 #include "svm_onhyperv.h"
 
 MODULE_AUTHOR("Qumranet");
+MODULE_DESCRIPTION("KVM support for SVM (AMD-V) extensions");
 MODULE_LICENSE("GPL");
 
 #ifdef MODULE
@@ -1202,7 +1203,7 @@ static inline void init_vmcb_after_set_cpuid(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
 
-	if (guest_cpuid_is_intel(vcpu)) {
+	if (guest_cpuid_is_intel_compatible(vcpu)) {
 		/*
 		 * We must intercept SYSENTER_EIP and SYSENTER_ESP
 		 * accesses because the processor only stores 32 bits.
@@ -2890,12 +2891,12 @@ static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 		break;
 	case MSR_IA32_SYSENTER_EIP:
 		msr_info->data = (u32)svm->vmcb01.ptr->save.sysenter_eip;
-		if (guest_cpuid_is_intel(vcpu))
+		if (guest_cpuid_is_intel_compatible(vcpu))
 			msr_info->data |= (u64)svm->sysenter_eip_hi << 32;
 		break;
 	case MSR_IA32_SYSENTER_ESP:
 		msr_info->data = svm->vmcb01.ptr->save.sysenter_esp;
-		if (guest_cpuid_is_intel(vcpu))
+		if (guest_cpuid_is_intel_compatible(vcpu))
 			msr_info->data |= (u64)svm->sysenter_esp_hi << 32;
 		break;
 	case MSR_TSC_AUX:
@@ -3122,11 +3123,11 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
 		 * 32 bit part of these msrs to support Intel's
 		 * implementation of SYSENTER/SYSEXIT.
 		 */
-		svm->sysenter_eip_hi = guest_cpuid_is_intel(vcpu) ? (data >> 32) : 0;
+		svm->sysenter_eip_hi = guest_cpuid_is_intel_compatible(vcpu) ? (data >> 32) : 0;
 		break;
 	case MSR_IA32_SYSENTER_ESP:
 		svm->vmcb01.ptr->save.sysenter_esp = (u32)data;
-		svm->sysenter_esp_hi = guest_cpuid_is_intel(vcpu) ? (data >> 32) : 0;
+		svm->sysenter_esp_hi = guest_cpuid_is_intel_compatible(vcpu) ? (data >> 32) : 0;
 		break;
 	case MSR_TSC_AUX:
 		/*
@@ -4387,11 +4388,11 @@ static void svm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
 	kvm_governed_feature_check_and_set(vcpu, X86_FEATURE_LBRV);
 
 	/*
-	 * Intercept VMLOAD if the vCPU mode is Intel in order to emulate that
+	 * Intercept VMLOAD if the vCPU model is Intel in order to emulate that
 	 * VMLOAD drops bits 63:32 of SYSENTER (ignoring the fact that exposing
 	 * SVM on Intel is bonkers and extremely unlikely to work).
 	 */
-	if (!guest_cpuid_is_intel(vcpu))
+	if (!guest_cpuid_is_intel_compatible(vcpu))
 		kvm_governed_feature_check_and_set(vcpu, X86_FEATURE_V_VMSAVE_VMLOAD);
 
 	kvm_governed_feature_check_and_set(vcpu, X86_FEATURE_PAUSEFILTER);

arch/x86/kvm/svm/svm.h

@@ -668,7 +668,7 @@ extern struct kvm_x86_nested_ops svm_nested_ops;
 /* avic.c */
 #define AVIC_REQUIRED_APICV_INHIBITS			\
 (							\
-	BIT(APICV_INHIBIT_REASON_DISABLE) |		\
+	BIT(APICV_INHIBIT_REASON_DISABLED) |		\
 	BIT(APICV_INHIBIT_REASON_ABSENT) |		\
 	BIT(APICV_INHIBIT_REASON_HYPERV) |		\
 	BIT(APICV_INHIBIT_REASON_NESTED) |		\

arch/x86/kvm/trace.h

@@ -1375,6 +1375,10 @@ TRACE_EVENT(kvm_hv_stimer_cleanup,
 		  __entry->vcpu_id, __entry->timer_index)
 );
 
+#define kvm_print_apicv_inhibit_reasons(inhibits)	\
+	(inhibits), (inhibits) ? " " : "",		\
+	(inhibits) ? __print_flags(inhibits, "|", APICV_INHIBIT_REASONS) : ""
+
 TRACE_EVENT(kvm_apicv_inhibit_changed,
 	    TP_PROTO(int reason, bool set, unsigned long inhibits),
 	    TP_ARGS(reason, set, inhibits),
@@ -1391,9 +1395,10 @@ TRACE_EVENT(kvm_apicv_inhibit_changed,
 		__entry->inhibits = inhibits;
 	),
 
-	TP_printk("%s reason=%u, inhibits=0x%lx",
+	TP_printk("%s reason=%u, inhibits=0x%lx%s%s",
 		  __entry->set ? "set" : "cleared",
-		  __entry->reason, __entry->inhibits)
+		  __entry->reason,
+		  kvm_print_apicv_inhibit_reasons(__entry->inhibits))
 );
 
 TRACE_EVENT(kvm_apicv_accept_irq,

arch/x86/kvm/vmx/main.c

@@ -8,7 +8,7 @@
 #include "posted_intr.h"
 
 #define VMX_REQUIRED_APICV_INHIBITS				\
-	(BIT(APICV_INHIBIT_REASON_DISABLE)|			\
+	(BIT(APICV_INHIBIT_REASON_DISABLED) |			\
 	 BIT(APICV_INHIBIT_REASON_ABSENT) |			\
 	 BIT(APICV_INHIBIT_REASON_HYPERV) |			\
 	 BIT(APICV_INHIBIT_REASON_BLOCKIRQ) |			\

arch/x86/kvm/vmx/nested.c

@@ -2425,7 +2425,7 @@ static void prepare_vmcs02_early(struct vcpu_vmx *vmx, struct loaded_vmcs *vmcs0
 	if (cpu_has_load_ia32_efer()) {
 		if (guest_efer & EFER_LMA)
 			exec_control |= VM_ENTRY_IA32E_MODE;
-		if (guest_efer != host_efer)
+		if (guest_efer != kvm_host.efer)
 			exec_control |= VM_ENTRY_LOAD_IA32_EFER;
 	}
 	vm_entry_controls_set(vmx, exec_control);
@@ -2438,7 +2438,7 @@ static void prepare_vmcs02_early(struct vcpu_vmx *vmx, struct loaded_vmcs *vmcs0
 	 * bits may be modified by vmx_set_efer() in prepare_vmcs02().
 	 */
 	exec_control = __vm_exit_controls_get(vmcs01);
-	if (cpu_has_load_ia32_efer() && guest_efer != host_efer)
+	if (cpu_has_load_ia32_efer() && guest_efer != kvm_host.efer)
 		exec_control |= VM_EXIT_LOAD_IA32_EFER;
 	else
 		exec_control &= ~VM_EXIT_LOAD_IA32_EFER;
@@ -4665,7 +4665,7 @@ static inline u64 nested_vmx_get_vmcs01_guest_efer(struct vcpu_vmx *vmx)
 		return vmcs_read64(GUEST_IA32_EFER);
 
 	if (cpu_has_load_ia32_efer())
-		return host_efer;
+		return kvm_host.efer;
 
 	for (i = 0; i < vmx->msr_autoload.guest.nr; ++i) {
 		if (vmx->msr_autoload.guest.val[i].index == MSR_EFER)
@@ -4676,7 +4676,7 @@ static inline u64 nested_vmx_get_vmcs01_guest_efer(struct vcpu_vmx *vmx)
 	if (efer_msr)
 		return efer_msr->data;
 
-	return host_efer;
+	return kvm_host.efer;
 }
 
 static void nested_vmx_restore_host_state(struct kvm_vcpu *vcpu)

arch/x86/kvm/vmx/vmx.c

@@ -74,6 +74,7 @@
 #include "posted_intr.h"
 
 MODULE_AUTHOR("Qumranet");
+MODULE_DESCRIPTION("KVM support for VMX (Intel VT-x) extensions");
 MODULE_LICENSE("GPL");
 
 #ifdef MODULE
@@ -259,7 +260,7 @@ static int vmx_setup_l1d_flush(enum vmx_l1d_flush_state l1tf)
 		return 0;
 	}
 
-	if (host_arch_capabilities & ARCH_CAP_SKIP_VMENTRY_L1DFLUSH) {
+	if (kvm_host.arch_capabilities & ARCH_CAP_SKIP_VMENTRY_L1DFLUSH) {
 		l1tf_vmx_mitigation = VMENTER_L1D_FLUSH_NOT_REQUIRED;
 		return 0;
 	}
@@ -404,7 +405,7 @@ static void vmx_update_fb_clear_dis(struct kvm_vcpu *vcpu, struct vcpu_vmx *vmx)
 	 * and VM-Exit.
 	 */
 	vmx->disable_fb_clear = !cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF) &&
-				(host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
+				(kvm_host.arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
 				!boot_cpu_has_bug(X86_BUG_MDS) &&
 				!boot_cpu_has_bug(X86_BUG_TAA);
 
@@ -1123,12 +1124,12 @@ static bool update_transition_efer(struct vcpu_vmx *vmx)
 	 * atomically, since it's faster than switching it manually.
 	 */
 	if (cpu_has_load_ia32_efer() ||
-	    (enable_ept && ((vmx->vcpu.arch.efer ^ host_efer) & EFER_NX))) {
+	    (enable_ept && ((vmx->vcpu.arch.efer ^ kvm_host.efer) & EFER_NX))) {
 		if (!(guest_efer & EFER_LMA))
 			guest_efer &= ~EFER_LME;
-		if (guest_efer != host_efer)
+		if (guest_efer != kvm_host.efer)
 			add_atomic_switch_msr(vmx, MSR_EFER,
-					      guest_efer, host_efer, false);
+					      guest_efer, kvm_host.efer, false);
 		else
 			clear_atomic_switch_msr(vmx, MSR_EFER);
 		return false;
@@ -1141,7 +1142,7 @@ static bool update_transition_efer(struct vcpu_vmx *vmx)
 	clear_atomic_switch_msr(vmx, MSR_EFER);
 
 	guest_efer &= ~ignore_bits;
-	guest_efer |= host_efer & ignore_bits;
+	guest_efer |= kvm_host.efer & ignore_bits;
 
 	vmx->guest_uret_msrs[i].data = guest_efer;
 	vmx->guest_uret_msrs[i].mask = ~ignore_bits;
@@ -4392,7 +4393,7 @@ void vmx_set_constant_host_state(struct vcpu_vmx *vmx)
 	}
 
 	if (cpu_has_load_ia32_efer())
-		vmcs_write64(HOST_IA32_EFER, host_efer);
+		vmcs_write64(HOST_IA32_EFER, kvm_host.efer);
 }
 
 void set_cr4_guest_host_mask(struct vcpu_vmx *vmx)
@@ -8394,18 +8395,16 @@ static void __init vmx_setup_me_spte_mask(void)
 	u64 me_mask = 0;
 
 	/*
-	 * kvm_get_shadow_phys_bits() returns shadow_phys_bits.  Use
-	 * the former to avoid exposing shadow_phys_bits.
-	 *
 	 * On pre-MKTME system, boot_cpu_data.x86_phys_bits equals to
-	 * shadow_phys_bits.  On MKTME and/or TDX capable systems,
+	 * kvm_host.maxphyaddr.  On MKTME and/or TDX capable systems,
 	 * boot_cpu_data.x86_phys_bits holds the actual physical address
-	 * w/o the KeyID bits, and shadow_phys_bits equals to MAXPHYADDR
+	 * w/o the KeyID bits, and kvm_host.maxphyaddr equals to
-	 * reported by CPUID.  Those bits between are KeyID bits.
+	 * MAXPHYADDR reported by CPUID.  Those bits between are KeyID bits.
 	 */
-	if (boot_cpu_data.x86_phys_bits != kvm_get_shadow_phys_bits())
+	if (boot_cpu_data.x86_phys_bits != kvm_host.maxphyaddr)
 		me_mask = rsvd_bits(boot_cpu_data.x86_phys_bits,
-			kvm_get_shadow_phys_bits() - 1);
+				    kvm_host.maxphyaddr - 1);
+
 	/*
 	 * Unlike SME, host kernel doesn't support setting up any
 	 * MKTME KeyID on Intel platforms.  No memory encryption

arch/x86/kvm/vmx/vmx.h

@@ -727,7 +727,7 @@ static inline bool vmx_need_pf_intercept(struct kvm_vcpu *vcpu)
 		return true;
 
 	return allow_smaller_maxphyaddr &&
-	       cpuid_maxphyaddr(vcpu) < kvm_get_shadow_phys_bits();
+	       cpuid_maxphyaddr(vcpu) < kvm_host.maxphyaddr;
 }
 
 static inline bool is_unrestricted_guest(struct kvm_vcpu *vcpu)

arch/x86/kvm/vmx/x86_ops.h

@@ -46,7 +46,6 @@ bool vmx_apic_init_signal_blocked(struct kvm_vcpu *vcpu);
 void vmx_migrate_timers(struct kvm_vcpu *vcpu);
 void vmx_set_virtual_apic_mode(struct kvm_vcpu *vcpu);
 void vmx_apicv_pre_state_restore(struct kvm_vcpu *vcpu);
-bool vmx_check_apicv_inhibit_reasons(enum kvm_apicv_inhibit reason);
 void vmx_hwapic_irr_update(struct kvm_vcpu *vcpu, int max_irr);
 void vmx_hwapic_isr_update(int max_isr);
 bool vmx_guest_apic_has_interrupt(struct kvm_vcpu *vcpu);
@@ -111,7 +110,6 @@ u64 vmx_get_l2_tsc_offset(struct kvm_vcpu *vcpu);
 u64 vmx_get_l2_tsc_multiplier(struct kvm_vcpu *vcpu);
 void vmx_write_tsc_offset(struct kvm_vcpu *vcpu);
 void vmx_write_tsc_multiplier(struct kvm_vcpu *vcpu);
-void vmx_request_immediate_exit(struct kvm_vcpu *vcpu);
 void vmx_update_cpu_dirty_logging(struct kvm_vcpu *vcpu);
 #ifdef CONFIG_X86_64
 int vmx_set_hv_timer(struct kvm_vcpu *vcpu, u64 guest_deadline_tsc,

arch/x86/kvm/x86.c

@@ -100,6 +100,9 @@
 struct kvm_caps kvm_caps __read_mostly;
 EXPORT_SYMBOL_GPL(kvm_caps);
 
+struct kvm_host_values kvm_host __read_mostly;
+EXPORT_SYMBOL_GPL(kvm_host);
+
 #define  ERR_PTR_USR(e)  ((void __user *)ERR_PTR(e))
 
 #define emul_to_vcpu(ctxt) \
@@ -220,21 +223,12 @@ static struct kvm_user_return_msrs __percpu *user_return_msrs;
 				| XFEATURE_MASK_BNDCSR | XFEATURE_MASK_AVX512 \
 				| XFEATURE_MASK_PKRU | XFEATURE_MASK_XTILE)
 
-u64 __read_mostly host_efer;
-EXPORT_SYMBOL_GPL(host_efer);
-
 bool __read_mostly allow_smaller_maxphyaddr = 0;
 EXPORT_SYMBOL_GPL(allow_smaller_maxphyaddr);
 
 bool __read_mostly enable_apicv = true;
 EXPORT_SYMBOL_GPL(enable_apicv);
 
-u64 __read_mostly host_xss;
-EXPORT_SYMBOL_GPL(host_xss);
-
-u64 __read_mostly host_arch_capabilities;
-EXPORT_SYMBOL_GPL(host_arch_capabilities);
-
 const struct _kvm_stats_desc kvm_vm_stats_desc[] = {
 	KVM_GENERIC_VM_STATS(),
 	STATS_DESC_COUNTER(VM, mmu_shadow_zapped),
@@ -308,8 +302,6 @@ const struct kvm_stats_header kvm_vcpu_stats_header = {
 		       sizeof(kvm_vcpu_stats_desc),
 };
 
-u64 __read_mostly host_xcr0;
-
 static struct kmem_cache *x86_emulator_cache;
 
 /*
@@ -1016,11 +1008,11 @@ void kvm_load_guest_xsave_state(struct kvm_vcpu *vcpu)
 
 	if (kvm_is_cr4_bit_set(vcpu, X86_CR4_OSXSAVE)) {
 
-		if (vcpu->arch.xcr0 != host_xcr0)
+		if (vcpu->arch.xcr0 != kvm_host.xcr0)
 			xsetbv(XCR_XFEATURE_ENABLED_MASK, vcpu->arch.xcr0);
 
 		if (guest_can_use(vcpu, X86_FEATURE_XSAVES) &&
-		    vcpu->arch.ia32_xss != host_xss)
+		    vcpu->arch.ia32_xss != kvm_host.xss)
 			wrmsrl(MSR_IA32_XSS, vcpu->arch.ia32_xss);
 	}
 
@@ -1047,12 +1039,12 @@ void kvm_load_host_xsave_state(struct kvm_vcpu *vcpu)
 
 	if (kvm_is_cr4_bit_set(vcpu, X86_CR4_OSXSAVE)) {
 
-		if (vcpu->arch.xcr0 != host_xcr0)
+		if (vcpu->arch.xcr0 != kvm_host.xcr0)
-			xsetbv(XCR_XFEATURE_ENABLED_MASK, host_xcr0);
+			xsetbv(XCR_XFEATURE_ENABLED_MASK, kvm_host.xcr0);
 
 		if (guest_can_use(vcpu, X86_FEATURE_XSAVES) &&
-		    vcpu->arch.ia32_xss != host_xss)
+		    vcpu->arch.ia32_xss != kvm_host.xss)
-			wrmsrl(MSR_IA32_XSS, host_xss);
+			wrmsrl(MSR_IA32_XSS, kvm_host.xss);
 	}
 
 }
@@ -1619,7 +1611,7 @@ static bool kvm_is_immutable_feature_msr(u32 msr)
 
 static u64 kvm_get_arch_capabilities(void)
 {
-	u64 data = host_arch_capabilities & KVM_SUPPORTED_ARCH_CAP;
+	u64 data = kvm_host.arch_capabilities & KVM_SUPPORTED_ARCH_CAP;
 
 	/*
 	 * If nx_huge_pages is enabled, KVM's shadow paging will ensure that
@@ -1877,11 +1869,11 @@ static int __kvm_set_msr(struct kvm_vcpu *vcpu, u32 index, u64 data,
 		 * incomplete and conflicting architectural behavior.  Current
 		 * AMD CPUs completely ignore bits 63:32, i.e. they aren't
 		 * reserved and always read as zeros.  Enforce Intel's reserved
-		 * bits check if and only if the guest CPU is Intel, and clear
+		 * bits check if the guest CPU is Intel compatible, otherwise
-		 * the bits in all other cases.  This ensures cross-vendor
+		 * clear the bits.  This ensures cross-vendor migration will
-		 * migration will provide consistent behavior for the guest.
+		 * provide consistent behavior for the guest.
 		 */
-		if (guest_cpuid_is_intel(vcpu) && (data >> 32) != 0)
+		if (guest_cpuid_is_intel_compatible(vcpu) && (data >> 32) != 0)
 			return 1;
 
 		data = (u32)data;
@@ -4703,11 +4695,15 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 	case KVM_CAP_VM_DISABLE_NX_HUGE_PAGES:
 	case KVM_CAP_IRQFD_RESAMPLE:
 	case KVM_CAP_MEMORY_FAULT_INFO:
+	case KVM_CAP_X86_GUEST_MODE:
 		r = 1;
 		break;
 	case KVM_CAP_PRE_FAULT_MEMORY:
 		r = tdp_enabled;
 		break;
+	case KVM_CAP_X86_APIC_BUS_CYCLES_NS:
+		r = APIC_BUS_CYCLE_NS_DEFAULT;
+		break;
 	case KVM_CAP_EXIT_HYPERCALL:
 		r = KVM_EXIT_HYPERCALL_VALID_MASK;
 		break;
@@ -5891,8 +5887,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 		if (!lapic_in_kernel(vcpu))
 			goto out;
-		u.lapic = kzalloc(sizeof(struct kvm_lapic_state),
+		u.lapic = kzalloc(sizeof(struct kvm_lapic_state), GFP_KERNEL);
-				GFP_KERNEL_ACCOUNT);
 
 		r = -ENOMEM;
 		if (!u.lapic)
@@ -6085,7 +6080,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (vcpu->arch.guest_fpu.uabi_size > sizeof(struct kvm_xsave))
 			break;
 
-		u.xsave = kzalloc(sizeof(struct kvm_xsave), GFP_KERNEL_ACCOUNT);
+		u.xsave = kzalloc(sizeof(struct kvm_xsave), GFP_KERNEL);
 		r = -ENOMEM;
 		if (!u.xsave)
 			break;
@@ -6116,7 +6111,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	case KVM_GET_XSAVE2: {
 		int size = vcpu->arch.guest_fpu.uabi_size;
 
-		u.xsave = kzalloc(size, GFP_KERNEL_ACCOUNT);
+		u.xsave = kzalloc(size, GFP_KERNEL);
 		r = -ENOMEM;
 		if (!u.xsave)
 			break;
@@ -6134,7 +6129,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 
 	case KVM_GET_XCRS: {
-		u.xcrs = kzalloc(sizeof(struct kvm_xcrs), GFP_KERNEL_ACCOUNT);
+		u.xcrs = kzalloc(sizeof(struct kvm_xcrs), GFP_KERNEL);
 		r = -ENOMEM;
 		if (!u.xcrs)
 			break;
@@ -6756,6 +6751,30 @@ split_irqchip_unlock:
 		}
 		mutex_unlock(&kvm->lock);
 		break;
+	case KVM_CAP_X86_APIC_BUS_CYCLES_NS: {
+		u64 bus_cycle_ns = cap->args[0];
+		u64 unused;
+
+		/*
+		 * Guard against overflow in tmict_to_ns(). 128 is the highest
+		 * divide value that can be programmed in APIC_TDCR.
+		 */
+		r = -EINVAL;
+		if (!bus_cycle_ns ||
+		    check_mul_overflow((u64)U32_MAX * 128, bus_cycle_ns, &unused))
+			break;
+
+		r = 0;
+		mutex_lock(&kvm->lock);
+		if (!irqchip_in_kernel(kvm))
+			r = -ENXIO;
+		else if (kvm->created_vcpus)
+			r = -EINVAL;
+		else
+			kvm->arch.apic_bus_cycle_ns = bus_cycle_ns;
+		mutex_unlock(&kvm->lock);
+		break;
+	}
 	default:
 		r = -EINVAL;
 		break;
@@ -8535,6 +8554,11 @@ static bool emulator_guest_has_rdpid(struct x86_emulate_ctxt *ctxt)
 	return guest_cpuid_has(emul_to_vcpu(ctxt), X86_FEATURE_RDPID);
 }
 
+static bool emulator_guest_cpuid_is_intel_compatible(struct x86_emulate_ctxt *ctxt)
+{
+	return guest_cpuid_is_intel_compatible(emul_to_vcpu(ctxt));
+}
+
 static ulong emulator_read_gpr(struct x86_emulate_ctxt *ctxt, unsigned reg)
 {
 	return kvm_register_read_raw(emul_to_vcpu(ctxt), reg);
@@ -8633,6 +8657,7 @@ static const struct x86_emulate_ops emulate_ops = {
 	.guest_has_movbe     = emulator_guest_has_movbe,
 	.guest_has_fxsr      = emulator_guest_has_fxsr,
 	.guest_has_rdpid     = emulator_guest_has_rdpid,
+	.guest_cpuid_is_intel_compatible = emulator_guest_cpuid_is_intel_compatible,
 	.set_nmi_mask        = emulator_set_nmi_mask,
 	.is_smm              = emulator_is_smm,
 	.is_guest_mode       = emulator_is_guest_mode,
@@ -9014,19 +9039,17 @@ EXPORT_SYMBOL_GPL(kvm_skip_emulated_instruction);
 
 static bool kvm_is_code_breakpoint_inhibited(struct kvm_vcpu *vcpu)
 {
-	u32 shadow;
-
 	if (kvm_get_rflags(vcpu) & X86_EFLAGS_RF)
 		return true;
 
 	/*
-	 * Intel CPUs inhibit code #DBs when MOV/POP SS blocking is active,
+	 * Intel compatible CPUs inhibit code #DBs when MOV/POP SS blocking is
-	 * but AMD CPUs do not.  MOV/POP SS blocking is rare, check that first
+	 * active, but AMD compatible CPUs do not.
-	 * to avoid the relatively expensive CPUID lookup.
 	 */
-	shadow = static_call(kvm_x86_get_interrupt_shadow)(vcpu);
+	if (!guest_cpuid_is_intel_compatible(vcpu))
-	return (shadow & KVM_X86_SHADOW_INT_MOV_SS) &&
+		return false;
-	       guest_cpuid_is_intel(vcpu);
+
+	return static_call(kvm_x86_get_interrupt_shadow)(vcpu) & KVM_X86_SHADOW_INT_MOV_SS;
 }
 
 static bool kvm_vcpu_check_code_breakpoint(struct kvm_vcpu *vcpu,
@@ -9786,19 +9809,19 @@ int kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
 	kvm_caps.supported_mce_cap = MCG_CTL_P | MCG_SER_P;
 
 	if (boot_cpu_has(X86_FEATURE_XSAVE)) {
-		host_xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
+		kvm_host.xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
-		kvm_caps.supported_xcr0 = host_xcr0 & KVM_SUPPORTED_XCR0;
+		kvm_caps.supported_xcr0 = kvm_host.xcr0 & KVM_SUPPORTED_XCR0;
 	}
 
-	rdmsrl_safe(MSR_EFER, &host_efer);
+	rdmsrl_safe(MSR_EFER, &kvm_host.efer);
 
 	if (boot_cpu_has(X86_FEATURE_XSAVES))
-		rdmsrl(MSR_IA32_XSS, host_xss);
+		rdmsrl(MSR_IA32_XSS, kvm_host.xss);
 
 	kvm_init_pmu_capability(ops->pmu_ops);
 
 	if (boot_cpu_has(X86_FEATURE_ARCH_CAPABILITIES))
-		rdmsrl(MSR_IA32_ARCH_CAPABILITIES, host_arch_capabilities);
+		rdmsrl(MSR_IA32_ARCH_CAPABILITIES, kvm_host.arch_capabilities);
 
 	r = ops->hardware_setup();
 	if (r != 0)
@@ -10023,6 +10046,10 @@ EXPORT_SYMBOL_GPL(kvm_vcpu_apicv_activated);
 static void set_or_clear_apicv_inhibit(unsigned long *inhibits,
 				       enum kvm_apicv_inhibit reason, bool set)
 {
+	const struct trace_print_flags apicv_inhibits[] = { APICV_INHIBIT_REASONS };
+
+	BUILD_BUG_ON(ARRAY_SIZE(apicv_inhibits) != NR_APICV_INHIBIT_REASONS);
+
 	if (set)
 		__set_bit(reason, inhibits);
 	else
@@ -10034,7 +10061,7 @@ static void set_or_clear_apicv_inhibit(unsigned long *inhibits,
 static void kvm_apicv_init(struct kvm *kvm)
 {
 	enum kvm_apicv_inhibit reason = enable_apicv ? APICV_INHIBIT_REASON_ABSENT :
-						       APICV_INHIBIT_REASON_DISABLE;
+						       APICV_INHIBIT_REASON_DISABLED;
 
 	set_or_clear_apicv_inhibit(&kvm->arch.apicv_inhibit_reasons, reason, true);
 
@@ -10255,6 +10282,8 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu)
 
 	if (is_smm(vcpu))
 		kvm_run->flags |= KVM_RUN_X86_SMM;
+	if (is_guest_mode(vcpu))
+		kvm_run->flags |= KVM_RUN_X86_GUEST_MODE;
 }
 
 static void update_cr8_intercept(struct kvm_vcpu *vcpu)
@@ -12629,6 +12658,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	raw_spin_unlock_irqrestore(&kvm->arch.tsc_write_lock, flags);
 
 	kvm->arch.default_tsc_khz = max_tsc_khz ? : tsc_khz;
+	kvm->arch.apic_bus_cycle_ns = APIC_BUS_CYCLE_NS_DEFAULT;
 	kvm->arch.guest_can_read_msr_platform_info = true;
 	kvm->arch.enable_pmu = enable_pmu;
 

arch/x86/kvm/x86.h

@@ -33,6 +33,20 @@ struct kvm_caps {
 	u64 supported_perf_cap;
 };
 
+struct kvm_host_values {
+	/*
+	 * The host's raw MAXPHYADDR, i.e. the number of non-reserved physical
+	 * address bits irrespective of features that repurpose legal bits,
+	 * e.g. MKTME.
+	 */
+	u8 maxphyaddr;
+
+	u64 efer;
+	u64 xcr0;
+	u64 xss;
+	u64 arch_capabilities;
+};
+
 void kvm_spurious_fault(void);
 
 #define KVM_NESTED_VMENTER_CONSISTENCY_CHECK(consistency_check)		\
@@ -325,11 +339,8 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa,
 			    int emulation_type, void *insn, int insn_len);
 fastpath_t handle_fastpath_set_msr_irqoff(struct kvm_vcpu *vcpu);
 
-extern u64 host_xcr0;
-extern u64 host_xss;
-extern u64 host_arch_capabilities;
-
 extern struct kvm_caps kvm_caps;
+extern struct kvm_host_values kvm_host;
 
 extern bool enable_pmu;
 

include/uapi/linux/kvm.h

@@ -931,6 +931,8 @@ struct kvm_enable_cap {
 #define KVM_CAP_GUEST_MEMFD 234
 #define KVM_CAP_VM_TYPES 235
 #define KVM_CAP_PRE_FAULT_MEMORY 236
+#define KVM_CAP_X86_APIC_BUS_CYCLES_NS 237
+#define KVM_CAP_X86_GUEST_MODE 238
 
 struct kvm_irq_routing_irqchip {
 	__u32 irqchip;

tools/testing/selftests/kvm/Makefile

@@ -112,6 +112,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/vmx_invalid_nested_guest_state
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_set_nested_state_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_tsc_adjust_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_nested_tsc_scaling_test
+TEST_GEN_PROGS_x86_64 += x86_64/apic_bus_clock_test
 TEST_GEN_PROGS_x86_64 += x86_64/xapic_ipi_test
 TEST_GEN_PROGS_x86_64 += x86_64/xapic_state_test
 TEST_GEN_PROGS_x86_64 += x86_64/xcr0_cpuid_test

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

@@ -60,6 +60,14 @@
 #define		APIC_VECTOR_MASK	0x000FF
 #define	APIC_ICR2	0x310
 #define		SET_APIC_DEST_FIELD(x)	((x) << 24)
+#define APIC_LVTT	0x320
+#define		APIC_LVT_TIMER_ONESHOT		(0 << 17)
+#define		APIC_LVT_TIMER_PERIODIC		(1 << 17)
+#define		APIC_LVT_TIMER_TSCDEADLINE	(2 << 17)
+#define		APIC_LVT_MASKED			(1 << 16)
+#define	APIC_TMICT	0x380
+#define	APIC_TMCCT	0x390
+#define	APIC_TDCR	0x3E0
 
 void apic_disable(void);
 void xapic_enable(void);

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

@@ -23,6 +23,7 @@
 
 extern bool host_cpu_is_intel;
 extern bool host_cpu_is_amd;
+extern uint64_t guest_tsc_khz;
 
 /* Forced emulation prefix, used to invoke the emulator unconditionally. */
 #define KVM_FEP "ud2; .byte 'k', 'v', 'm';"
@@ -816,6 +817,23 @@ static inline void cpu_relax(void)
 	asm volatile("rep; nop" ::: "memory");
 }
 
+static inline void udelay(unsigned long usec)
+{
+	uint64_t start, now, cycles;
+
+	GUEST_ASSERT(guest_tsc_khz);
+	cycles = guest_tsc_khz / 1000 * usec;
+
+	/*
+	 * Deliberately don't PAUSE, a.k.a. cpu_relax(), so that the delay is
+	 * as accurate as possible, e.g. doesn't trigger PAUSE-Loop VM-Exits.
+	 */
+	start = rdtsc();
+	do {
+		now = rdtsc();
+	} while (now - start < cycles);
+}
+
 #define ud2()			\
 	__asm__ __volatile__(	\
 		"ud2\n"	\

tools/testing/selftests/kvm/lib/x86_64/processor.c

@@ -25,6 +25,7 @@ vm_vaddr_t exception_handlers;
 bool host_cpu_is_amd;
 bool host_cpu_is_intel;
 bool is_forced_emulation_enabled;
+uint64_t guest_tsc_khz;
 
 static void regs_dump(FILE *stream, struct kvm_regs *regs, uint8_t indent)
 {
@@ -616,6 +617,11 @@ void assert_on_unhandled_exception(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vm_post_create(struct kvm_vm *vm)
 {
+	int r;
+
+	TEST_ASSERT(kvm_has_cap(KVM_CAP_GET_TSC_KHZ),
+		    "Require KVM_GET_TSC_KHZ to provide udelay() to guest.");
+
 	vm_create_irqchip(vm);
 	vm_init_descriptor_tables(vm);
 
@@ -628,6 +634,11 @@ void kvm_arch_vm_post_create(struct kvm_vm *vm)
 
 		vm_sev_ioctl(vm, KVM_SEV_INIT2, &init);
 	}
+
+	r = __vm_ioctl(vm, KVM_GET_TSC_KHZ, NULL);
+	TEST_ASSERT(r > 0, "KVM_GET_TSC_KHZ did not provide a valid TSC frequency.");
+	guest_tsc_khz = r;
+	sync_global_to_guest(vm, guest_tsc_khz);
 }
 
 void vcpu_arch_set_entry_point(struct kvm_vcpu *vcpu, void *guest_code)

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

@@ -0,0 +1,194 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (c) 2024 Intel Corporation
+ *
+ * Verify KVM correctly emulates the APIC bus frequency when the VMM configures
+ * the frequency via KVM_CAP_X86_APIC_BUS_CYCLES_NS.  Start the APIC timer by
+ * programming TMICT (timer initial count) to the largest value possible (so
+ * that the timer will not expire during the test).  Then, after an arbitrary
+ * amount of time has elapsed, verify TMCCT (timer current count) is within 1%
+ * of the expected value based on the time elapsed, the APIC bus frequency, and
+ * the programmed TDCR (timer divide configuration register).
+ */
+
+#include "apic.h"
+#include "test_util.h"
+
+/*
+ * Possible TDCR values with matching divide count. Used to modify APIC
+ * timer frequency.
+ */
+static const struct {
+	const uint32_t tdcr;
+	const uint32_t divide_count;
+} tdcrs[] = {
+	{0x0, 2},
+	{0x1, 4},
+	{0x2, 8},
+	{0x3, 16},
+	{0x8, 32},
+	{0x9, 64},
+	{0xa, 128},
+	{0xb, 1},
+};
+
+static bool is_x2apic;
+
+static void apic_enable(void)
+{
+	if (is_x2apic)
+		x2apic_enable();
+	else
+		xapic_enable();
+}
+
+static uint32_t apic_read_reg(unsigned int reg)
+{
+	return is_x2apic ? x2apic_read_reg(reg) : xapic_read_reg(reg);
+}
+
+static void apic_write_reg(unsigned int reg, uint32_t val)
+{
+	if (is_x2apic)
+		x2apic_write_reg(reg, val);
+	else
+		xapic_write_reg(reg, val);
+}
+
+static void apic_guest_code(uint64_t apic_hz, uint64_t delay_ms)
+{
+	uint64_t tsc_hz = guest_tsc_khz * 1000;
+	const uint32_t tmict = ~0u;
+	uint64_t tsc0, tsc1, freq;
+	uint32_t tmcct;
+	int i;
+
+	apic_enable();
+
+	/*
+	 * Setup one-shot timer.  The vector does not matter because the
+	 * interrupt should not fire.
+	 */
+	apic_write_reg(APIC_LVTT, APIC_LVT_TIMER_ONESHOT | APIC_LVT_MASKED);
+
+	for (i = 0; i < ARRAY_SIZE(tdcrs); i++) {
+		apic_write_reg(APIC_TDCR, tdcrs[i].tdcr);
+		apic_write_reg(APIC_TMICT, tmict);
+
+		tsc0 = rdtsc();
+		udelay(delay_ms * 1000);
+		tmcct = apic_read_reg(APIC_TMCCT);
+		tsc1 = rdtsc();
+
+		/*
+		 * Stop the timer _after_ reading the current, final count, as
+		 * writing the initial counter also modifies the current count.
+		 */
+		apic_write_reg(APIC_TMICT, 0);
+
+		freq = (tmict - tmcct) * tdcrs[i].divide_count * tsc_hz / (tsc1 - tsc0);
+		/* Check if measured frequency is within 5% of configured frequency. */
+		__GUEST_ASSERT(freq < apic_hz * 105 / 100 && freq > apic_hz * 95 / 100,
+			       "Frequency = %lu (wanted %lu - %lu), bus = %lu, div = %u, tsc = %lu",
+			       freq, apic_hz * 95 / 100, apic_hz * 105 / 100,
+			       apic_hz, tdcrs[i].divide_count, tsc_hz);
+	}
+
+	GUEST_DONE();
+}
+
+static void test_apic_bus_clock(struct kvm_vcpu *vcpu)
+{
+	bool done = false;
+	struct ucall uc;
+
+	while (!done) {
+		vcpu_run(vcpu);
+
+		TEST_ASSERT_KVM_EXIT_REASON(vcpu, KVM_EXIT_IO);
+
+		switch (get_ucall(vcpu, &uc)) {
+		case UCALL_DONE:
+			done = true;
+			break;
+		case UCALL_ABORT:
+			REPORT_GUEST_ASSERT(uc);
+			break;
+		default:
+			TEST_FAIL("Unknown ucall %lu", uc.cmd);
+			break;
+		}
+	}
+}
+
+static void run_apic_bus_clock_test(uint64_t apic_hz, uint64_t delay_ms,
+				    bool x2apic)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	int ret;
+
+	is_x2apic = x2apic;
+
+	vm = vm_create(1);
+
+	sync_global_to_guest(vm, is_x2apic);
+
+	vm_enable_cap(vm, KVM_CAP_X86_APIC_BUS_CYCLES_NS,
+		      NSEC_PER_SEC / apic_hz);
+
+	vcpu = vm_vcpu_add(vm, 0, apic_guest_code);
+	vcpu_args_set(vcpu, 2, apic_hz, delay_ms);
+
+	ret = __vm_enable_cap(vm, KVM_CAP_X86_APIC_BUS_CYCLES_NS,
+			      NSEC_PER_SEC / apic_hz);
+	TEST_ASSERT(ret < 0 && errno == EINVAL,
+		    "Setting of APIC bus frequency after vCPU is created should fail.");
+
+	if (!is_x2apic)
+		virt_pg_map(vm, APIC_DEFAULT_GPA, APIC_DEFAULT_GPA);
+
+	test_apic_bus_clock(vcpu);
+	kvm_vm_free(vm);
+}
+
+static void help(char *name)
+{
+	puts("");
+	printf("usage: %s [-h] [-d delay] [-f APIC bus freq]\n", name);
+	puts("");
+	printf("-d: Delay (in msec) guest uses to measure APIC bus frequency.\n");
+	printf("-f: The APIC bus frequency (in MHz) to be configured for the guest.\n");
+	puts("");
+}
+
+int main(int argc, char *argv[])
+{
+	/*
+	 * Arbitrarilty default to 25MHz for the APIC bus frequency, which is
+	 * different enough from the default 1GHz to be interesting.
+	 */
+	uint64_t apic_hz = 25 * 1000 * 1000;
+	uint64_t delay_ms = 100;
+	int opt;
+
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_X86_APIC_BUS_CYCLES_NS));
+
+	while ((opt = getopt(argc, argv, "d:f:h")) != -1) {
+		switch (opt) {
+		case 'f':
+			apic_hz = atoi_positive("APIC bus frequency", optarg) * 1000 * 1000;
+			break;
+		case 'd':
+			delay_ms = atoi_positive("Delay in milliseconds", optarg);
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			exit(KSFT_SKIP);
+		}
+	}
+
+	run_apic_bus_clock_test(apic_hz, delay_ms, false);
+	run_apic_bus_clock_test(apic_hz, delay_ms, true);
+}