diff options
Diffstat (limited to 'arch/x86')
34 files changed, 1502 insertions, 1111 deletions
diff --git a/arch/x86/include/asm/cpufeatures.h b/arch/x86/include/asm/cpufeatures.h index 353b054812de..cb8ca46213be 100644 --- a/arch/x86/include/asm/cpufeatures.h +++ b/arch/x86/include/asm/cpufeatures.h @@ -226,10 +226,9 @@ /* Virtualization flags: Linux defined, word 8 */ #define X86_FEATURE_TPR_SHADOW ( 8*32+ 0) /* Intel TPR Shadow */ -#define X86_FEATURE_VNMI ( 8*32+ 1) /* Intel Virtual NMI */ -#define X86_FEATURE_FLEXPRIORITY ( 8*32+ 2) /* Intel FlexPriority */ -#define X86_FEATURE_EPT ( 8*32+ 3) /* Intel Extended Page Table */ -#define X86_FEATURE_VPID ( 8*32+ 4) /* Intel Virtual Processor ID */ +#define X86_FEATURE_FLEXPRIORITY ( 8*32+ 1) /* Intel FlexPriority */ +#define X86_FEATURE_EPT ( 8*32+ 2) /* Intel Extended Page Table */ +#define X86_FEATURE_VPID ( 8*32+ 3) /* Intel Virtual Processor ID */ #define X86_FEATURE_VMMCALL ( 8*32+15) /* Prefer VMMCALL to VMCALL */ #define X86_FEATURE_XENPV ( 8*32+16) /* "" Xen paravirtual guest */ @@ -338,6 +337,7 @@ #define X86_FEATURE_VIRT_SSBD (13*32+25) /* Virtualized Speculative Store Bypass Disable */ #define X86_FEATURE_AMD_SSB_NO (13*32+26) /* "" Speculative Store Bypass is fixed in hardware. */ #define X86_FEATURE_CPPC (13*32+27) /* Collaborative Processor Performance Control */ +#define X86_FEATURE_AMD_PSFD (13*32+28) /* "" Predictive Store Forwarding Disable */ #define X86_FEATURE_BTC_NO (13*32+29) /* "" Not vulnerable to Branch Type Confusion */ #define X86_FEATURE_BRS (13*32+31) /* Branch Sampling available */ @@ -370,6 +370,7 @@ #define X86_FEATURE_VGIF (15*32+16) /* Virtual GIF */ #define X86_FEATURE_X2AVIC (15*32+18) /* Virtual x2apic */ #define X86_FEATURE_V_SPEC_CTRL (15*32+20) /* Virtual SPEC_CTRL */ +#define X86_FEATURE_VNMI (15*32+25) /* Virtual NMI */ #define X86_FEATURE_SVME_ADDR_CHK (15*32+28) /* "" SVME addr check */ /* Intel-defined CPU features, CPUID level 0x00000007:0 (ECX), word 16 */ diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h index 8dc345cc6318..13bc212cd4bc 100644 --- a/arch/x86/include/asm/kvm-x86-ops.h +++ b/arch/x86/include/asm/kvm-x86-ops.h @@ -54,8 +54,8 @@ KVM_X86_OP(set_rflags) KVM_X86_OP(get_if_flag) KVM_X86_OP(flush_tlb_all) KVM_X86_OP(flush_tlb_current) -KVM_X86_OP_OPTIONAL(tlb_remote_flush) -KVM_X86_OP_OPTIONAL(tlb_remote_flush_with_range) +KVM_X86_OP_OPTIONAL(flush_remote_tlbs) +KVM_X86_OP_OPTIONAL(flush_remote_tlbs_range) KVM_X86_OP(flush_tlb_gva) KVM_X86_OP(flush_tlb_guest) KVM_X86_OP(vcpu_pre_run) @@ -68,6 +68,8 @@ KVM_X86_OP(get_interrupt_shadow) KVM_X86_OP(patch_hypercall) KVM_X86_OP(inject_irq) KVM_X86_OP(inject_nmi) +KVM_X86_OP_OPTIONAL_RET0(is_vnmi_pending) +KVM_X86_OP_OPTIONAL_RET0(set_vnmi_pending) KVM_X86_OP(inject_exception) KVM_X86_OP(cancel_injection) KVM_X86_OP(interrupt_allowed) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 808c292ad3f4..fb9d1f2d6136 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -420,6 +420,10 @@ struct kvm_mmu_root_info { #define KVM_MMU_NUM_PREV_ROOTS 3 +#define KVM_MMU_ROOT_CURRENT BIT(0) +#define KVM_MMU_ROOT_PREVIOUS(i) BIT(1+i) +#define KVM_MMU_ROOTS_ALL (BIT(1 + KVM_MMU_NUM_PREV_ROOTS) - 1) + #define KVM_HAVE_MMU_RWLOCK struct kvm_mmu_page; @@ -439,9 +443,8 @@ struct kvm_mmu { gpa_t (*gva_to_gpa)(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, gpa_t gva_or_gpa, u64 access, struct x86_exception *exception); - int (*sync_page)(struct kvm_vcpu *vcpu, - struct kvm_mmu_page *sp); - void (*invlpg)(struct kvm_vcpu *vcpu, gva_t gva, hpa_t root_hpa); + int (*sync_spte)(struct kvm_vcpu *vcpu, + struct kvm_mmu_page *sp, int i); struct kvm_mmu_root_info root; union kvm_cpu_role cpu_role; union kvm_mmu_page_role root_role; @@ -479,11 +482,6 @@ struct kvm_mmu { u64 pdptrs[4]; /* pae */ }; -struct kvm_tlb_range { - u64 start_gfn; - u64 pages; -}; - enum pmc_type { KVM_PMC_GP = 0, KVM_PMC_FIXED, @@ -515,6 +513,7 @@ struct kvm_pmc { #define MSR_ARCH_PERFMON_FIXED_CTR_MAX (MSR_ARCH_PERFMON_FIXED_CTR0 + KVM_PMC_MAX_FIXED - 1) #define KVM_AMD_PMC_MAX_GENERIC 6 struct kvm_pmu { + u8 version; unsigned nr_arch_gp_counters; unsigned nr_arch_fixed_counters; unsigned available_event_types; @@ -527,7 +526,6 @@ struct kvm_pmu { u64 global_ovf_ctrl_mask; u64 reserved_bits; u64 raw_event_mask; - u8 version; struct kvm_pmc gp_counters[KVM_INTEL_PMC_MAX_GENERIC]; struct kvm_pmc fixed_counters[KVM_PMC_MAX_FIXED]; struct irq_work irq_work; @@ -876,7 +874,8 @@ struct kvm_vcpu_arch { u64 tsc_scaling_ratio; /* current scaling ratio */ atomic_t nmi_queued; /* unprocessed asynchronous NMIs */ - unsigned nmi_pending; /* NMI queued after currently running handler */ + /* Number of NMIs pending injection, not including hardware vNMIs. */ + unsigned int nmi_pending; bool nmi_injected; /* Trying to inject an NMI this entry */ bool smi_pending; /* SMI queued after currently running handler */ u8 handling_intr_from_guest; @@ -947,23 +946,6 @@ struct kvm_vcpu_arch { u64 msr_kvm_poll_control; - /* - * Indicates the guest is trying to write a gfn that contains one or - * more of the PTEs used to translate the write itself, i.e. the access - * is changing its own translation in the guest page tables. KVM exits - * to userspace if emulation of the faulting instruction fails and this - * flag is set, as KVM cannot make forward progress. - * - * If emulation fails for a write to guest page tables, KVM unprotects - * (zaps) the shadow page for the target gfn and resumes the guest to - * retry the non-emulatable instruction (on hardware). Unprotecting the - * gfn doesn't allow forward progress for a self-changing access because - * doing so also zaps the translation for the gfn, i.e. retrying the - * instruction will hit a !PRESENT fault, which results in a new shadow - * page and sends KVM back to square one. - */ - bool write_fault_to_shadow_pgtable; - /* set at EPT violation at this point */ unsigned long exit_qualification; @@ -1602,9 +1584,9 @@ struct kvm_x86_ops { void (*flush_tlb_all)(struct kvm_vcpu *vcpu); void (*flush_tlb_current)(struct kvm_vcpu *vcpu); - int (*tlb_remote_flush)(struct kvm *kvm); - int (*tlb_remote_flush_with_range)(struct kvm *kvm, - struct kvm_tlb_range *range); + int (*flush_remote_tlbs)(struct kvm *kvm); + int (*flush_remote_tlbs_range)(struct kvm *kvm, gfn_t gfn, + gfn_t nr_pages); /* * Flush any TLB entries associated with the given GVA. @@ -1638,6 +1620,13 @@ struct kvm_x86_ops { int (*nmi_allowed)(struct kvm_vcpu *vcpu, bool for_injection); bool (*get_nmi_mask)(struct kvm_vcpu *vcpu); void (*set_nmi_mask)(struct kvm_vcpu *vcpu, bool masked); + /* Whether or not a virtual NMI is pending in hardware. */ + bool (*is_vnmi_pending)(struct kvm_vcpu *vcpu); + /* + * Attempt to pend a virtual NMI in harware. Returns %true on success + * to allow using static_call_ret0 as the fallback. + */ + bool (*set_vnmi_pending)(struct kvm_vcpu *vcpu); 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); @@ -1808,8 +1797,8 @@ void kvm_arch_free_vm(struct kvm *kvm); #define __KVM_HAVE_ARCH_FLUSH_REMOTE_TLB static inline int kvm_arch_flush_remote_tlb(struct kvm *kvm) { - if (kvm_x86_ops.tlb_remote_flush && - !static_call(kvm_x86_tlb_remote_flush)(kvm)) + if (kvm_x86_ops.flush_remote_tlbs && + !static_call(kvm_x86_flush_remote_tlbs)(kvm)) return 0; else return -ENOTSUPP; @@ -1907,6 +1896,25 @@ u64 vcpu_tsc_khz(struct kvm_vcpu *vcpu); * EMULTYPE_COMPLETE_USER_EXIT - Set when the emulator should update interruptibility * state and inject single-step #DBs after skipping * an instruction (after completing userspace I/O). + * + * EMULTYPE_WRITE_PF_TO_SP - Set when emulating an intercepted page fault that + * is attempting to write a gfn that contains one or + * more of the PTEs used to translate the write itself, + * and the owning page table is being shadowed by KVM. + * If emulation of the faulting instruction fails and + * this flag is set, KVM will exit to userspace instead + * of retrying emulation as KVM cannot make forward + * progress. + * + * If emulation fails for a write to guest page tables, + * KVM unprotects (zaps) the shadow page for the target + * gfn and resumes the guest to retry the non-emulatable + * instruction (on hardware). Unprotecting the gfn + * doesn't allow forward progress for a self-changing + * access because doing so also zaps the translation for + * the gfn, i.e. retrying the instruction will hit a + * !PRESENT fault, which results in a new shadow page + * and sends KVM back to square one. */ #define EMULTYPE_NO_DECODE (1 << 0) #define EMULTYPE_TRAP_UD (1 << 1) @@ -1916,6 +1924,7 @@ u64 vcpu_tsc_khz(struct kvm_vcpu *vcpu); #define EMULTYPE_VMWARE_GP (1 << 5) #define EMULTYPE_PF (1 << 6) #define EMULTYPE_COMPLETE_USER_EXIT (1 << 7) +#define EMULTYPE_WRITE_PF_TO_SP (1 << 8) int kvm_emulate_instruction(struct kvm_vcpu *vcpu, int emulation_type); int kvm_emulate_instruction_from_buffer(struct kvm_vcpu *vcpu, @@ -1994,14 +2003,11 @@ static inline int __kvm_irq_line_state(unsigned long *irq_state, return !!(*irq_state); } -#define KVM_MMU_ROOT_CURRENT BIT(0) -#define KVM_MMU_ROOT_PREVIOUS(i) BIT(1+i) -#define KVM_MMU_ROOTS_ALL (~0UL) - int kvm_pic_set_irq(struct kvm_pic *pic, int irq, int irq_source_id, int level); void kvm_pic_clear_all(struct kvm_pic *pic, int irq_source_id); void kvm_inject_nmi(struct kvm_vcpu *vcpu); +int kvm_get_nr_pending_nmis(struct kvm_vcpu *vcpu); void kvm_update_dr7(struct kvm_vcpu *vcpu); @@ -2041,8 +2047,8 @@ int kvm_emulate_hypercall(struct kvm_vcpu *vcpu); int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa, u64 error_code, void *insn, int insn_len); void kvm_mmu_invlpg(struct kvm_vcpu *vcpu, gva_t gva); -void kvm_mmu_invalidate_gva(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, - gva_t gva, hpa_t root_hpa); +void kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, + u64 addr, unsigned long roots); void kvm_mmu_invpcid_gva(struct kvm_vcpu *vcpu, gva_t gva, unsigned long pcid); void kvm_mmu_new_pgd(struct kvm_vcpu *vcpu, gpa_t new_pgd); @@ -2204,4 +2210,11 @@ int memslot_rmap_alloc(struct kvm_memory_slot *slot, unsigned long npages); KVM_X86_QUIRK_FIX_HYPERCALL_INSN | \ KVM_X86_QUIRK_MWAIT_NEVER_UD_FAULTS) +/* + * KVM previously used a u32 field in kvm_run to indicate the hypercall was + * initiated from long mode. KVM now sets bit 0 to indicate long mode, but the + * remaining 31 lower bits must be 0 to preserve ABI. + */ +#define KVM_EXIT_HYPERCALL_MBZ GENMASK_ULL(31, 1) + #endif /* _ASM_X86_KVM_HOST_H */ diff --git a/arch/x86/include/asm/svm.h b/arch/x86/include/asm/svm.h index 770dcf75eaa9..e7c7379d6ac7 100644 --- a/arch/x86/include/asm/svm.h +++ b/arch/x86/include/asm/svm.h @@ -183,6 +183,12 @@ struct __attribute__ ((__packed__)) vmcb_control_area { #define V_GIF_SHIFT 9 #define V_GIF_MASK (1 << V_GIF_SHIFT) +#define V_NMI_PENDING_SHIFT 11 +#define V_NMI_PENDING_MASK (1 << V_NMI_PENDING_SHIFT) + +#define V_NMI_BLOCKING_SHIFT 12 +#define V_NMI_BLOCKING_MASK (1 << V_NMI_BLOCKING_SHIFT) + #define V_INTR_PRIO_SHIFT 16 #define V_INTR_PRIO_MASK (0x0f << V_INTR_PRIO_SHIFT) @@ -197,6 +203,9 @@ struct __attribute__ ((__packed__)) vmcb_control_area { #define V_GIF_ENABLE_SHIFT 25 #define V_GIF_ENABLE_MASK (1 << V_GIF_ENABLE_SHIFT) +#define V_NMI_ENABLE_SHIFT 26 +#define V_NMI_ENABLE_MASK (1 << V_NMI_ENABLE_SHIFT) + #define AVIC_ENABLE_SHIFT 31 #define AVIC_ENABLE_MASK (1 << AVIC_ENABLE_SHIFT) @@ -278,7 +287,6 @@ static_assert((AVIC_MAX_PHYSICAL_ID & AVIC_PHYSICAL_MAX_INDEX_MASK) == AVIC_MAX_ static_assert((X2AVIC_MAX_PHYSICAL_ID & AVIC_PHYSICAL_MAX_INDEX_MASK) == X2AVIC_MAX_PHYSICAL_ID); #define AVIC_HPA_MASK ~((0xFFFULL << 52) | 0xFFF) -#define VMCB_AVIC_APIC_BAR_MASK 0xFFFFFFFFFF000ULL struct vmcb_seg { diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h index 7f467fe05d42..1a6a1f987949 100644 --- a/arch/x86/include/uapi/asm/kvm.h +++ b/arch/x86/include/uapi/asm/kvm.h @@ -559,4 +559,7 @@ struct kvm_pmu_event_filter { #define KVM_VCPU_TSC_CTRL 0 /* control group for the timestamp counter (TSC) */ #define KVM_VCPU_TSC_OFFSET 0 /* attribute for the TSC offset */ +/* x86-specific KVM_EXIT_HYPERCALL flags. */ +#define KVM_EXIT_HYPERCALL_LONG_MODE BIT(0) + #endif /* _ASM_X86_KVM_H */ diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c index 599aebec2d52..123bf8b97a4b 100644 --- a/arch/x86/kvm/cpuid.c +++ b/arch/x86/kvm/cpuid.c @@ -60,12 +60,6 @@ u32 xstate_required_size(u64 xstate_bv, bool compacted) return ret; } -/* - * This one is tied to SSB in the user API, and not - * visible in /proc/cpuinfo. - */ -#define KVM_X86_FEATURE_AMD_PSFD (13*32+28) /* Predictive Store Forwarding Disable */ - #define F feature_bit /* Scattered Flag - For features that are scattered by cpufeatures.h. */ @@ -266,7 +260,7 @@ static void __kvm_update_cpuid_runtime(struct kvm_vcpu *vcpu, struct kvm_cpuid_e /* Update OSXSAVE bit */ if (boot_cpu_has(X86_FEATURE_XSAVE)) cpuid_entry_change(best, X86_FEATURE_OSXSAVE, - kvm_read_cr4_bits(vcpu, X86_CR4_OSXSAVE)); + kvm_is_cr4_bit_set(vcpu, X86_CR4_OSXSAVE)); cpuid_entry_change(best, X86_FEATURE_APIC, vcpu->arch.apic_base & MSR_IA32_APICBASE_ENABLE); @@ -275,7 +269,7 @@ static void __kvm_update_cpuid_runtime(struct kvm_vcpu *vcpu, struct kvm_cpuid_e best = cpuid_entry2_find(entries, nent, 7, 0); if (best && boot_cpu_has(X86_FEATURE_PKU) && best->function == 0x7) cpuid_entry_change(best, X86_FEATURE_OSPKE, - kvm_read_cr4_bits(vcpu, X86_CR4_PKE)); + kvm_is_cr4_bit_set(vcpu, X86_CR4_PKE)); best = cpuid_entry2_find(entries, nent, 0xD, 0); if (best) @@ -420,7 +414,7 @@ static int kvm_set_cpuid(struct kvm_vcpu *vcpu, struct kvm_cpuid_entry2 *e2, * KVM_SET_CPUID{,2} again. To support this legacy behavior, check * whether the supplied CPUID data is equal to what's already set. */ - if (vcpu->arch.last_vmentry_cpu != -1) { + if (kvm_vcpu_has_run(vcpu)) { r = kvm_cpuid_check_equal(vcpu, e2, nent); if (r) return r; @@ -653,7 +647,7 @@ void kvm_set_cpu_caps(void) F(SPEC_CTRL_SSBD) | F(ARCH_CAPABILITIES) | F(INTEL_STIBP) | F(MD_CLEAR) | F(AVX512_VP2INTERSECT) | F(FSRM) | F(SERIALIZE) | F(TSXLDTRK) | F(AVX512_FP16) | - F(AMX_TILE) | F(AMX_INT8) | F(AMX_BF16) + F(AMX_TILE) | F(AMX_INT8) | F(AMX_BF16) | F(FLUSH_L1D) ); /* TSC_ADJUST and ARCH_CAPABILITIES are emulated in software. */ @@ -715,7 +709,7 @@ void kvm_set_cpu_caps(void) F(CLZERO) | F(XSAVEERPTR) | F(WBNOINVD) | F(AMD_IBPB) | F(AMD_IBRS) | F(AMD_SSBD) | F(VIRT_SSBD) | F(AMD_SSB_NO) | F(AMD_STIBP) | F(AMD_STIBP_ALWAYS_ON) | - __feature_bit(KVM_X86_FEATURE_AMD_PSFD) + F(AMD_PSFD) ); /* @@ -1002,7 +996,7 @@ static inline int __do_cpuid_func(struct kvm_cpuid_array *array, u32 function) entry->eax = entry->ebx = entry->ecx = 0; break; case 0xd: { - u64 permitted_xcr0 = kvm_caps.supported_xcr0 & xstate_get_guest_group_perm(); + u64 permitted_xcr0 = kvm_get_filtered_xcr0(); u64 permitted_xss = kvm_caps.supported_xss; entry->eax &= permitted_xcr0; diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index a20bec931764..936a397a08cd 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -1640,6 +1640,14 @@ static int __load_segment_descriptor(struct x86_emulate_ctxt *ctxt, goto exception; break; case VCPU_SREG_CS: + /* + * KVM uses "none" when loading CS as part of emulating Real + * Mode exceptions and IRET (handled above). In all other + * cases, loading CS without a control transfer is a KVM bug. + */ + if (WARN_ON_ONCE(transfer == X86_TRANSFER_NONE)) + goto exception; + if (!(seg_desc.type & 8)) goto exception; diff --git a/arch/x86/kvm/kvm_cache_regs.h b/arch/x86/kvm/kvm_cache_regs.h index 4c91f626c058..75eae9c4998a 100644 --- a/arch/x86/kvm/kvm_cache_regs.h +++ b/arch/x86/kvm/kvm_cache_regs.h @@ -4,7 +4,7 @@ #include <linux/kvm_host.h> -#define KVM_POSSIBLE_CR0_GUEST_BITS X86_CR0_TS +#define KVM_POSSIBLE_CR0_GUEST_BITS (X86_CR0_TS | X86_CR0_WP) #define KVM_POSSIBLE_CR4_GUEST_BITS \ (X86_CR4_PVI | X86_CR4_DE | X86_CR4_PCE | X86_CR4_OSFXSR \ | X86_CR4_OSXMMEXCPT | X86_CR4_PGE | X86_CR4_TSD | X86_CR4_FSGSBASE) @@ -157,6 +157,14 @@ static inline ulong kvm_read_cr0_bits(struct kvm_vcpu *vcpu, ulong mask) return vcpu->arch.cr0 & mask; } +static __always_inline bool kvm_is_cr0_bit_set(struct kvm_vcpu *vcpu, + unsigned long cr0_bit) +{ + BUILD_BUG_ON(!is_power_of_2(cr0_bit)); + + return !!kvm_read_cr0_bits(vcpu, cr0_bit); +} + static inline ulong kvm_read_cr0(struct kvm_vcpu *vcpu) { return kvm_read_cr0_bits(vcpu, ~0UL); @@ -171,6 +179,14 @@ static inline ulong kvm_read_cr4_bits(struct kvm_vcpu *vcpu, ulong mask) return vcpu->arch.cr4 & mask; } +static __always_inline bool kvm_is_cr4_bit_set(struct kvm_vcpu *vcpu, + unsigned long cr4_bit) +{ + BUILD_BUG_ON(!is_power_of_2(cr4_bit)); + + return !!kvm_read_cr4_bits(vcpu, cr4_bit); +} + static inline ulong kvm_read_cr3(struct kvm_vcpu *vcpu) { if (!kvm_register_is_available(vcpu, VCPU_EXREG_CR3)) diff --git a/arch/x86/kvm/kvm_onhyperv.c b/arch/x86/kvm/kvm_onhyperv.c index 482d6639ef88..ded0bd688c65 100644 --- a/arch/x86/kvm/kvm_onhyperv.c +++ b/arch/x86/kvm/kvm_onhyperv.c @@ -10,17 +10,22 @@ #include "hyperv.h" #include "kvm_onhyperv.h" +struct kvm_hv_tlb_range { + u64 start_gfn; + u64 pages; +}; + static int kvm_fill_hv_flush_list_func(struct hv_guest_mapping_flush_list *flush, void *data) { - struct kvm_tlb_range *range = data; + struct kvm_hv_tlb_range *range = data; return hyperv_fill_flush_guest_mapping_list(flush, range->start_gfn, range->pages); } static inline int hv_remote_flush_root_tdp(hpa_t root_tdp, - struct kvm_tlb_range *range) + struct kvm_hv_tlb_range *range) { if (range) return hyperv_flush_guest_mapping_range(root_tdp, @@ -29,8 +34,8 @@ static inline int hv_remote_flush_root_tdp(hpa_t root_tdp, return hyperv_flush_guest_mapping(root_tdp); } -int hv_remote_flush_tlb_with_range(struct kvm *kvm, - struct kvm_tlb_range *range) +static int __hv_flush_remote_tlbs_range(struct kvm *kvm, + struct kvm_hv_tlb_range *range) { struct kvm_arch *kvm_arch = &kvm->arch; struct kvm_vcpu *vcpu; @@ -86,19 +91,29 @@ int hv_remote_flush_tlb_with_range(struct kvm *kvm, spin_unlock(&kvm_arch->hv_root_tdp_lock); return ret; } -EXPORT_SYMBOL_GPL(hv_remote_flush_tlb_with_range); -int hv_remote_flush_tlb(struct kvm *kvm) +int hv_flush_remote_tlbs_range(struct kvm *kvm, gfn_t start_gfn, gfn_t nr_pages) +{ + struct kvm_hv_tlb_range range = { + .start_gfn = start_gfn, + .pages = nr_pages, + }; + + return __hv_flush_remote_tlbs_range(kvm, &range); +} +EXPORT_SYMBOL_GPL(hv_flush_remote_tlbs_range); + +int hv_flush_remote_tlbs(struct kvm *kvm) { - return hv_remote_flush_tlb_with_range(kvm, NULL); + return __hv_flush_remote_tlbs_range(kvm, NULL); } -EXPORT_SYMBOL_GPL(hv_remote_flush_tlb); +EXPORT_SYMBOL_GPL(hv_flush_remote_tlbs); void hv_track_root_tdp(struct kvm_vcpu *vcpu, hpa_t root_tdp) { struct kvm_arch *kvm_arch = &vcpu->kvm->arch; - if (kvm_x86_ops.tlb_remote_flush == hv_remote_flush_tlb) { + if (kvm_x86_ops.flush_remote_tlbs == hv_flush_remote_tlbs) { spin_lock(&kvm_arch->hv_root_tdp_lock); vcpu->arch.hv_root_tdp = root_tdp; if (root_tdp != kvm_arch->hv_root_tdp) diff --git a/arch/x86/kvm/kvm_onhyperv.h b/arch/x86/kvm/kvm_onhyperv.h index 6272dabec02d..f9ca3e7432b2 100644 --- a/arch/x86/kvm/kvm_onhyperv.h +++ b/arch/x86/kvm/kvm_onhyperv.h @@ -7,12 +7,11 @@ #define __ARCH_X86_KVM_KVM_ONHYPERV_H__ #if IS_ENABLED(CONFIG_HYPERV) -int hv_remote_flush_tlb_with_range(struct kvm *kvm, - struct kvm_tlb_range *range); -int hv_remote_flush_tlb(struct kvm *kvm); +int hv_flush_remote_tlbs_range(struct kvm *kvm, gfn_t gfn, gfn_t nr_pages); +int hv_flush_remote_tlbs(struct kvm *kvm); void hv_track_root_tdp(struct kvm_vcpu *vcpu, hpa_t root_tdp); #else /* !CONFIG_HYPERV */ -static inline int hv_remote_flush_tlb(struct kvm *kvm) +static inline int hv_flush_remote_tlbs(struct kvm *kvm) { return -EOPNOTSUPP; } diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h index 168c46fd8dd1..92d5a1924fc1 100644 --- a/arch/x86/kvm/mmu.h +++ b/arch/x86/kvm/mmu.h @@ -113,6 +113,8 @@ void kvm_init_shadow_ept_mmu(struct kvm_vcpu *vcpu, bool execonly, bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu); int kvm_handle_page_fault(struct kvm_vcpu *vcpu, u64 error_code, u64 fault_address, char *insn, int insn_len); +void __kvm_mmu_refresh_passthrough_bits(struct kvm_vcpu *vcpu, + struct kvm_mmu *mmu); int kvm_mmu_load(struct kvm_vcpu *vcpu); void kvm_mmu_unload(struct kvm_vcpu *vcpu); @@ -132,7 +134,7 @@ static inline unsigned long kvm_get_pcid(struct kvm_vcpu *vcpu, gpa_t cr3) { BUILD_BUG_ON((X86_CR3_PCID_MASK & PAGE_MASK) != 0); - return kvm_read_cr4_bits(vcpu, X86_CR4_PCIDE) + return kvm_is_cr4_bit_set(vcpu, X86_CR4_PCIDE) ? cr3 & X86_CR3_PCID_MASK : 0; } @@ -153,6 +155,24 @@ static inline void kvm_mmu_load_pgd(struct kvm_vcpu *vcpu) vcpu->arch.mmu->root_role.level); } +static inline void kvm_mmu_refresh_passthrough_bits(struct kvm_vcpu *vcpu, + struct kvm_mmu *mmu) +{ + /* + * When EPT is enabled, KVM may passthrough CR0.WP to the guest, i.e. + * @mmu's snapshot of CR0.WP and thus all related paging metadata may + * be stale. Refresh CR0.WP and the metadata on-demand when checking + * for permission faults. Exempt nested MMUs, i.e. MMUs for shadowing + * nEPT and nNPT, as CR0.WP is ignored in both cases. Note, KVM does + * need to refresh nested_mmu, a.k.a. the walker used to translate L2 + * GVAs to GPAs, as that "MMU" needs to honor L2's CR0.WP. + */ + if (!tdp_enabled || mmu == &vcpu->arch.guest_mmu) + return; + + __kvm_mmu_refresh_passthrough_bits(vcpu, mmu); +} + /* * Check if a given access (described through the I/D, W/R and U/S bits of a * page fault error code pfec) causes a permission fault with the given PTE @@ -184,8 +204,12 @@ static inline u8 permission_fault(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, u64 implicit_access = access & PFERR_IMPLICIT_ACCESS; bool not_smap = ((rflags & X86_EFLAGS_AC) | implicit_access) == X86_EFLAGS_AC; int index = (pfec + (not_smap << PFERR_RSVD_BIT)) >> 1; - bool fault = (mmu->permissions[index] >> pte_access) & 1; u32 errcode = PFERR_PRESENT_MASK; + bool fault; + + kvm_mmu_refresh_passthrough_bits(vcpu, mmu); + + fault = (mmu->permissions[index] >> pte_access) & 1; WARN_ON(pfec & (PFERR_PK_MASK | PFERR_RSVD_MASK)); if (unlikely(mmu->pkru_mask)) { diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index c8ebe542c565..c8961f45e3b1 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -125,17 +125,31 @@ module_param(dbg, bool, 0644); #define PTE_LIST_EXT 14 /* - * Slight optimization of cacheline layout, by putting `more' and `spte_count' - * at the start; then accessing it will only use one single cacheline for - * either full (entries==PTE_LIST_EXT) case or entries<=6. + * struct pte_list_desc is the core data structure used to implement a custom + * list for tracking a set of related SPTEs, e.g. all the SPTEs that map a + * given GFN when used in the context of rmaps. Using a custom list allows KVM + * to optimize for the common case where many GFNs will have at most a handful + * of SPTEs pointing at them, i.e. allows packing multiple SPTEs into a small + * memory footprint, which in turn improves runtime performance by exploiting + * cache locality. + * + * A list is comprised of one or more pte_list_desc objects (descriptors). + * Each individual descriptor stores up to PTE_LIST_EXT SPTEs. If a descriptor + * is full and a new SPTEs needs to be added, a new descriptor is allocated and + * becomes the head of the list. This means that by definitions, all tail + * descriptors are full. + * + * Note, the meta data fields are deliberately placed at the start of the + * structure to optimize the cacheline layout; accessing the descriptor will + * touch only a single cacheline so long as @spte_count<=6 (or if only the + * descriptors metadata is accessed). */ struct pte_list_desc { struct pte_list_desc *more; - /* - * Stores number of entries stored in the pte_list_desc. No need to be - * u64 but just for easier alignment. When PTE_LIST_EXT, means full. - */ - u64 spte_count; + /* The number of PTEs stored in _this_ descriptor. */ + u32 spte_count; + /* The number of PTEs stored in all tails of this descriptor. */ + u32 tail_count; u64 *sptes[PTE_LIST_EXT]; }; @@ -242,32 +256,35 @@ static struct kvm_mmu_role_regs vcpu_to_role_regs(struct kvm_vcpu *vcpu) return regs; } -static inline bool kvm_available_flush_tlb_with_range(void) +static unsigned long get_guest_cr3(struct kvm_vcpu *vcpu) { - return kvm_x86_ops.tlb_remote_flush_with_range; + return kvm_read_cr3(vcpu); } -static void kvm_flush_remote_tlbs_with_range(struct kvm *kvm, - struct kvm_tlb_range *range) +static inline unsigned long kvm_mmu_get_guest_pgd(struct kvm_vcpu *vcpu, + struct kvm_mmu *mmu) { - int ret = -ENOTSUPP; - - if (range && kvm_x86_ops.tlb_remote_flush_with_range) - ret = static_call(kvm_x86_tlb_remote_flush_with_range)(kvm, range); + if (IS_ENABLED(CONFIG_RETPOLINE) && mmu->get_guest_pgd == get_guest_cr3) + return kvm_read_cr3(vcpu); - if (ret) - kvm_flush_remote_tlbs(kvm); + return mmu->get_guest_pgd(vcpu); } -void kvm_flush_remote_tlbs_with_address(struct kvm *kvm, - u64 start_gfn, u64 pages) +static inline bool kvm_available_flush_remote_tlbs_range(void) { - struct kvm_tlb_range range; + return kvm_x86_ops.flush_remote_tlbs_range; +} - range.start_gfn = start_gfn; - range.pages = pages; +void kvm_flush_remote_tlbs_range(struct kvm *kvm, gfn_t start_gfn, + gfn_t nr_pages) +{ + int ret = -EOPNOTSUPP; - kvm_flush_remote_tlbs_with_range(kvm, &range); + if (kvm_x86_ops.flush_remote_tlbs_range) + ret = static_call(kvm_x86_flush_remote_tlbs_range)(kvm, start_gfn, + nr_pages); + if (ret) + kvm_flush_remote_tlbs(kvm); } static gfn_t kvm_mmu_page_get_gfn(struct kvm_mmu_page *sp, int index); @@ -888,9 +905,9 @@ static void unaccount_nx_huge_page(struct kvm *kvm, struct kvm_mmu_page *sp) untrack_possible_nx_huge_page(kvm, sp); } -static struct kvm_memory_slot * -gfn_to_memslot_dirty_bitmap(struct kvm_vcpu *vcpu, gfn_t gfn, - bool no_dirty_log) +static struct kvm_memory_slot *gfn_to_memslot_dirty_bitmap(struct kvm_vcpu *vcpu, + gfn_t gfn, + bool no_dirty_log) { struct kvm_memory_slot *slot; @@ -929,53 +946,69 @@ static int pte_list_add(struct kvm_mmu_memory_cache *cache, u64 *spte, desc->sptes[0] = (u64 *)rmap_head->val; desc->sptes[1] = spte; desc->spte_count = 2; + desc->tail_count = 0; rmap_head->val = (unsigned long)desc | 1; ++count; } else { rmap_printk("%p %llx many->many\n", spte, *spte); desc = (struct pte_list_desc *)(rmap_head->val & ~1ul); - while (desc->spte_count == PTE_LIST_EXT) { - count += PTE_LIST_EXT; - if (!desc->more) { - desc->more = kvm_mmu_memory_cache_alloc(cache); - desc = desc->more; - desc->spte_count = 0; - break; - } - desc = desc->more; + count = desc->tail_count + desc->spte_count; + |