From mboxrd@z Thu Jan 1 00:00:00 1970 From: Steve Rutherford Subject: Re: [RFC PATCH 1/4] KVM: x86: Split the APIC from the rest of IRQCHIP. Date: Wed, 13 May 2015 15:10:01 -0700 Message-ID: <20150513221001.GA23867@google.com> References: <1431481652-27268-1-git-send-email-srutherford@google.com> <555303CC.5030507@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvm@vger.kernel.org, ahonig@google.com To: Paolo Bonzini Return-path: Received: from mail-ig0-f182.google.com ([209.85.213.182]:35308 "EHLO mail-ig0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933971AbbEMWKT (ORCPT ); Wed, 13 May 2015 18:10:19 -0400 Received: by igbyr2 with SMTP id yr2so151885152igb.0 for ; Wed, 13 May 2015 15:10:19 -0700 (PDT) Content-Disposition: inline In-Reply-To: <555303CC.5030507@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On Wed, May 13, 2015 at 09:57:00AM +0200, Paolo Bonzini wrote: > > > On 13/05/2015 03:47, Steve Rutherford wrote: > > First patch in a series which enables the relocation of the > > PIC/IOAPIC/PIT to userspace. > > > > Adds capability KVM_CAP_SPLIT_IRQCHIP and ioctl KVM_SPLIT_IRQCHIP. > > > > KVM_SPLIT_IRQCHIP enables the construction of LAPICs without the rest > > of the irqchip. > > The ioctl name in the code is KVM_CREATE_SPLIT_IRQCHIP, but it doesn't > seem to create a local APIC. > This enables the creation of LAPICs every time a VCPU is created, which I guess is different. I'll change the language to reflect that. > > Compile tested for x86. > > The capability is fine. However, instead of introducing a new ioctl, > you can use KVM_CAP_ENABLE_CAP_VM and enable the capability once > (per-VM) using KVM_ENABLE_CAP. The enabling code is basically the same > you have in your code. > ENABLE_CAP seems like a better way to go. I'll update this to use it instead of the IOCTL. Steve > Paolo > > > Signed-off-by: Steve Rutherford > > Suggested-by: Andrew Honig > > --- > > Documentation/virtual/kvm/api.txt | 15 ++++++++++++ > > arch/powerpc/kvm/irq.h | 5 ++++ > > arch/s390/kvm/irq.h | 4 ++++ > > arch/x86/include/asm/kvm_host.h | 2 ++ > > arch/x86/kvm/assigned-dev.c | 4 ++-- > > arch/x86/kvm/irq.c | 6 ++--- > > arch/x86/kvm/irq.h | 11 +++++++++ > > arch/x86/kvm/irq_comm.c | 7 ++++++ > > arch/x86/kvm/lapic.c | 13 +++++++---- > > arch/x86/kvm/mmu.c | 2 +- > > arch/x86/kvm/svm.c | 4 ++-- > > arch/x86/kvm/vmx.c | 12 +++++----- > > arch/x86/kvm/x86.c | 49 +++++++++++++++++++++++++++------------ > > include/kvm/arm_vgic.h | 1 + > > include/linux/kvm_host.h | 1 + > > include/uapi/linux/kvm.h | 3 +++ > > virt/kvm/irqchip.c | 2 +- > > 17 files changed, 106 insertions(+), 35 deletions(-) > > > > diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt > > index 6955444..0744b4e 100644 > > --- a/Documentation/virtual/kvm/api.txt > > +++ b/Documentation/virtual/kvm/api.txt > > @@ -2979,6 +2979,21 @@ len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 > > and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), > > which is the maximum number of possibly pending cpu-local interrupts. > > > > +4.96 KVM_SPLIT_IRQCHIP > > + > > +Capability: KVM_CAP_SPLIT_IRQCHIP > > +Architectures: x86 > > +Type: VM ioctl > > +Parameters: None > > +Returns: 0 on success, -1 on error > > + > > +Create a local apic for each processor in the kernel. This differs from > > +KVM_CREATE_IRQCHIP in that it only creates the local apic; it creates neither > > +the ioapic nor the pic in the kernel. Also, enables in kernel routing of > > +interrupt requests. Fails if VCPU has already been created, or if the irqchip is > > +already in the kernel. > > + > > + > > 5. The kvm_run structure > > ------------------------ > > > > diff --git a/arch/powerpc/kvm/irq.h b/arch/powerpc/kvm/irq.h > > index 5a9a10b..5e6fa06 100644 > > --- a/arch/powerpc/kvm/irq.h > > +++ b/arch/powerpc/kvm/irq.h > > @@ -17,4 +17,9 @@ static inline int irqchip_in_kernel(struct kvm *kvm) > > return ret; > > } > > > > +static inline int lapic_in_kernel(struct kvm *kvm) > > +{ > > + return irqchip_in_kernel(kvm); > > +} > > + > > #endif > > diff --git a/arch/s390/kvm/irq.h b/arch/s390/kvm/irq.h > > index d98e415..db876c3 100644 > > --- a/arch/s390/kvm/irq.h > > +++ b/arch/s390/kvm/irq.h > > @@ -19,4 +19,8 @@ static inline int irqchip_in_kernel(struct kvm *kvm) > > return 1; > > } > > > > +static inline int lapic_in_kernel(struct kvm *kvm) > > +{ > > + return irqchip_in_kernel(kvm); > > +} > > #endif > > diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h > > index bbb8f4e..3ddc134 100644 > > --- a/arch/x86/include/asm/kvm_host.h > > +++ b/arch/x86/include/asm/kvm_host.h > > @@ -638,6 +638,8 @@ struct kvm_arch { > > bool boot_vcpu_runs_old_kvmclock; > > > > u64 disabled_quirks; > > + > > + bool irqchip_split; > > }; > > > > struct kvm_vm_stat { > > diff --git a/arch/x86/kvm/assigned-dev.c b/arch/x86/kvm/assigned-dev.c > > index d090ecf..1237e92 100644 > > --- a/arch/x86/kvm/assigned-dev.c > > +++ b/arch/x86/kvm/assigned-dev.c > > @@ -291,7 +291,7 @@ static int kvm_deassign_irq(struct kvm *kvm, > > { > > unsigned long guest_irq_type, host_irq_type; > > > > - if (!irqchip_in_kernel(kvm)) > > + if (!lapic_in_kernel(kvm)) > > return -EINVAL; > > /* no irq assignment to deassign */ > > if (!assigned_dev->irq_requested_type) > > @@ -568,7 +568,7 @@ static int kvm_vm_ioctl_assign_irq(struct kvm *kvm, > > struct kvm_assigned_dev_kernel *match; > > unsigned long host_irq_type, guest_irq_type; > > > > - if (!irqchip_in_kernel(kvm)) > > + if (!lapic_in_kernel(kvm)) > > return r; > > > > mutex_lock(&kvm->lock); > > diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c > > index a1ec6a50..706e47a 100644 > > --- a/arch/x86/kvm/irq.c > > +++ b/arch/x86/kvm/irq.c > > @@ -57,7 +57,7 @@ static int kvm_cpu_has_extint(struct kvm_vcpu *v) > > */ > > int kvm_cpu_has_injectable_intr(struct kvm_vcpu *v) > > { > > - if (!irqchip_in_kernel(v->kvm)) > > + if (!lapic_in_kernel(v->kvm)) > > return v->arch.interrupt.pending; > > > > if (kvm_cpu_has_extint(v)) > > @@ -75,7 +75,7 @@ int kvm_cpu_has_injectable_intr(struct kvm_vcpu *v) > > */ > > int kvm_cpu_has_interrupt(struct kvm_vcpu *v) > > { > > - if (!irqchip_in_kernel(v->kvm)) > > + if (!lapic_in_kernel(v->kvm)) > > return v->arch.interrupt.pending; > > > > if (kvm_cpu_has_extint(v)) > > @@ -103,7 +103,7 @@ int kvm_cpu_get_interrupt(struct kvm_vcpu *v) > > { > > int vector; > > > > - if (!irqchip_in_kernel(v->kvm)) > > + if (!lapic_in_kernel(v->kvm)) > > return v->arch.interrupt.nr; > > > > vector = kvm_cpu_get_extint(v); > > diff --git a/arch/x86/kvm/irq.h b/arch/x86/kvm/irq.h > > index ad68c73..e46abf3 100644 > > --- a/arch/x86/kvm/irq.h > > +++ b/arch/x86/kvm/irq.h > > @@ -92,6 +92,17 @@ static inline int irqchip_in_kernel(struct kvm *kvm) > > return ret; > > } > > > > +static inline int irqchip_split(struct kvm *kvm) > > +{ > > + return kvm->arch.irqchip_split; > > +} > > + > > +static inline int lapic_in_kernel(struct kvm *kvm) > > +{ > > + return irqchip_split(kvm) || irqchip_in_kernel(kvm); > > +} > > + > > + > > void kvm_pic_reset(struct kvm_kpic_state *s); > > > > void kvm_inject_pending_timer_irqs(struct kvm_vcpu *vcpu); > > diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c > > index 9efff9e..f43c59a 100644 > > --- a/arch/x86/kvm/irq_comm.c > > +++ b/arch/x86/kvm/irq_comm.c > > @@ -328,3 +328,10 @@ int kvm_setup_default_irq_routing(struct kvm *kvm) > > return kvm_set_irq_routing(kvm, default_routing, > > ARRAY_SIZE(default_routing), 0); > > } > > + > > +static const struct kvm_irq_routing_entry empty_routing[] = {}; > > + > > +int kvm_setup_empty_irq_routing(struct kvm *kvm) > > +{ > > + return kvm_set_irq_routing(kvm, empty_routing, 0, 0); > > +} > > diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c > > index dc5b57b..bc392a6 100644 > > --- a/arch/x86/kvm/lapic.c > > +++ b/arch/x86/kvm/lapic.c > > @@ -209,7 +209,8 @@ out: > > if (old) > > kfree_rcu(old, rcu); > > > > - kvm_vcpu_request_scan_ioapic(kvm); > > + if (!irqchip_split(kvm)) > > + kvm_vcpu_request_scan_ioapic(kvm); > > } > > > > static inline void apic_set_spiv(struct kvm_lapic *apic, u32 val) > > @@ -1819,7 +1820,8 @@ void kvm_apic_post_state_restore(struct kvm_vcpu *vcpu, > > kvm_x86_ops->hwapic_isr_update(vcpu->kvm, > > apic_find_highest_isr(apic)); > > kvm_make_request(KVM_REQ_EVENT, vcpu); > > - kvm_rtc_eoi_tracking_restore_one(vcpu); > > + if (!irqchip_split(vcpu->kvm)) > > + kvm_rtc_eoi_tracking_restore_one(vcpu); > > } > > > > void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu) > > @@ -1902,7 +1904,8 @@ static void apic_sync_pv_eoi_to_guest(struct kvm_vcpu *vcpu, > > /* Cache not set: could be safe but we don't bother. */ > > apic->highest_isr_cache == -1 || > > /* Need EOI to update ioapic. */ > > - kvm_ioapic_handles_vector(vcpu->kvm, apic->highest_isr_cache)) { > > + kvm_ioapic_handles_vector(vcpu->kvm, apic->highest_isr_cache) || > > + irqchip_split(vcpu->kvm)) { > > /* > > * PV EOI was disabled by apic_sync_pv_eoi_from_guest > > * so we need not do anything here. > > @@ -1958,7 +1961,7 @@ int kvm_x2apic_msr_write(struct kvm_vcpu *vcpu, u32 msr, u64 data) > > struct kvm_lapic *apic = vcpu->arch.apic; > > u32 reg = (msr - APIC_BASE_MSR) << 4; > > > > - if (!irqchip_in_kernel(vcpu->kvm) || !apic_x2apic_mode(apic)) > > + if (!lapic_in_kernel(vcpu->kvm) || !apic_x2apic_mode(apic)) > > return 1; > > > > if (reg == APIC_ICR2) > > @@ -1975,7 +1978,7 @@ int kvm_x2apic_msr_read(struct kvm_vcpu *vcpu, u32 msr, u64 *data) > > struct kvm_lapic *apic = vcpu->arch.apic; > > u32 reg = (msr - APIC_BASE_MSR) << 4, low, high = 0; > > > > - if (!irqchip_in_kernel(vcpu->kvm) || !apic_x2apic_mode(apic)) > > + if (!lapic_in_kernel(vcpu->kvm) || !apic_x2apic_mode(apic)) > > return 1; > > > > if (reg == APIC_DFR || reg == APIC_ICR2) { > > diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c > > index b78e83f..e5bf6db 100644 > > --- a/arch/x86/kvm/mmu.c > > +++ b/arch/x86/kvm/mmu.c > > @@ -3465,7 +3465,7 @@ static int kvm_arch_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn) > > > > static bool can_do_async_pf(struct kvm_vcpu *vcpu) > > { > > - if (unlikely(!irqchip_in_kernel(vcpu->kvm) || > > + if (unlikely(!lapic_in_kernel(vcpu->kvm) || > > kvm_event_needs_reinjection(vcpu))) > > return false; > > > > diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c > > index 8954d7a..7ac9ec2 100644 > > --- a/arch/x86/kvm/svm.c > > +++ b/arch/x86/kvm/svm.c > > @@ -3054,7 +3054,7 @@ static int cr8_write_interception(struct vcpu_svm *svm) > > u8 cr8_prev = kvm_get_cr8(&svm->vcpu); > > /* instruction emulation calls kvm_set_cr8() */ > > r = cr_interception(svm); > > - if (irqchip_in_kernel(svm->vcpu.kvm)) > > + if (lapic_in_kernel(svm->vcpu.kvm)) > > return r; > > if (cr8_prev <= kvm_get_cr8(&svm->vcpu)) > > return r; > > @@ -3295,7 +3295,7 @@ static int interrupt_window_interception(struct vcpu_svm *svm) > > * If the user space waits to inject interrupts, exit as soon as > > * possible > > */ > > - if (!irqchip_in_kernel(svm->vcpu.kvm) && > > + if (!lapic_in_kernel(svm->vcpu.kvm) && > > kvm_run->request_interrupt_window && > > !kvm_cpu_has_interrupt(&svm->vcpu)) { > > kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN; > > diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c > > index bcb61b0..a53747f 100644 > > --- a/arch/x86/kvm/vmx.c > > +++ b/arch/x86/kvm/vmx.c > > @@ -948,7 +948,7 @@ static inline bool cpu_has_vmx_tpr_shadow(void) > > > > static inline bool vm_need_tpr_shadow(struct kvm *kvm) > > { > > - return (cpu_has_vmx_tpr_shadow()) && (irqchip_in_kernel(kvm)); > > + return (cpu_has_vmx_tpr_shadow()) && lapic_in_kernel(kvm); > > } > > > > static inline bool cpu_has_secondary_exec_ctrls(void) > > @@ -1064,7 +1064,7 @@ static inline bool cpu_has_vmx_ple(void) > > > > static inline bool vm_need_virtualize_apic_accesses(struct kvm *kvm) > > { > > - return flexpriority_enabled && irqchip_in_kernel(kvm); > > + return flexpriority_enabled && lapic_in_kernel(kvm); > > } > > > > static inline bool cpu_has_vmx_vpid(void) > > @@ -4341,7 +4341,7 @@ static void vmx_disable_intercept_msr_write_x2apic(u32 msr) > > > > static int vmx_vm_has_apicv(struct kvm *kvm) > > { > > - return enable_apicv && irqchip_in_kernel(kvm); > > + return enable_apicv && lapic_in_kernel(kvm); > > } > > > > static int vmx_complete_nested_posted_interrupt(struct kvm_vcpu *vcpu) > > @@ -5317,7 +5317,7 @@ static int handle_cr(struct kvm_vcpu *vcpu) > > u8 cr8 = (u8)val; > > err = kvm_set_cr8(vcpu, cr8); > > kvm_complete_insn_gp(vcpu, err); > > - if (irqchip_in_kernel(vcpu->kvm)) > > + if (lapic_in_kernel(vcpu->kvm)) > > return 1; > > if (cr8_prev <= cr8) > > return 1; > > @@ -5534,7 +5534,7 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu) > > * If the user space waits to inject interrupts, exit as soon as > > * possible > > */ > > - if (!irqchip_in_kernel(vcpu->kvm) && > > + if (!lapic_in_kernel(vcpu->kvm) && > > vcpu->run->request_interrupt_window && > > !kvm_cpu_has_interrupt(vcpu)) { > > vcpu->run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN; > > @@ -9419,7 +9419,7 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12) > > /* vmcs12's VM_ENTRY_LOAD_IA32_EFER and VM_ENTRY_IA32E_MODE are > > * emulated by vmx_set_efer(), below. > > */ > > - vm_entry_controls_init(vmx, > > + vm_entry_controls_init(vmx, > > (vmcs12->vm_entry_controls & ~VM_ENTRY_LOAD_IA32_EFER & > > ~VM_ENTRY_IA32E_MODE) | > > (vmcs_config.vmentry_ctrl & ~VM_ENTRY_IA32E_MODE)); > > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c > > index cde5d61..7505b39 100644 > > --- a/arch/x86/kvm/x86.c > > +++ b/arch/x86/kvm/x86.c > > @@ -782,7 +782,7 @@ int kvm_set_cr8(struct kvm_vcpu *vcpu, unsigned long cr8) > > { > > if (cr8 & CR8_RESERVED_BITS) > > return 1; > > - if (irqchip_in_kernel(vcpu->kvm)) > > + if (lapic_in_kernel(vcpu->kvm)) > > kvm_lapic_set_tpr(vcpu, cr8); > > else > > vcpu->arch.cr8 = cr8; > > @@ -792,7 +792,7 @@ EXPORT_SYMBOL_GPL(kvm_set_cr8); > > > > unsigned long kvm_get_cr8(struct kvm_vcpu *vcpu) > > { > > - if (irqchip_in_kernel(vcpu->kvm)) > > + if (lapic_in_kernel(vcpu->kvm)) > > return kvm_lapic_get_cr8(vcpu); > > else > > return vcpu->arch.cr8; > > @@ -2800,6 +2800,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) > > case KVM_CAP_TSC_DEADLINE_TIMER: > > case KVM_CAP_ENABLE_CAP_VM: > > case KVM_CAP_DISABLE_QUIRKS: > > + case KVM_CAP_SPLIT_IRQCHIP: > > #ifdef CONFIG_KVM_DEVICE_ASSIGNMENT > > case KVM_CAP_ASSIGN_DEV_IRQ: > > case KVM_CAP_PCI_2_3: > > @@ -3002,7 +3003,7 @@ static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, > > { > > if (irq->irq >= KVM_NR_INTERRUPTS) > > return -EINVAL; > > - if (irqchip_in_kernel(vcpu->kvm)) > > + if (lapic_in_kernel(vcpu->kvm)) > > return -ENXIO; > > > > kvm_queue_interrupt(vcpu, irq->irq, false); > > @@ -3480,7 +3481,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp, > > struct kvm_vapic_addr va; > > > > r = -EINVAL; > > - if (!irqchip_in_kernel(vcpu->kvm)) > > + if (!lapic_in_kernel(vcpu->kvm)) > > goto out; > > r = -EFAULT; > > if (copy_from_user(&va, argp, sizeof va)) > > @@ -3838,7 +3839,7 @@ int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm, struct kvm_dirty_log *log) > > int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_event, > > bool line_status) > > { > > - if (!irqchip_in_kernel(kvm)) > > + if (!lapic_in_kernel(kvm)) > > return -ENXIO; > > > > irq_event->status = kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID, > > @@ -4128,6 +4129,24 @@ long kvm_arch_vm_ioctl(struct file *filp, > > r = kvm_vm_ioctl_enable_cap(kvm, &cap); > > break; > > } > > + case KVM_CREATE_SPLIT_IRQCHIP: { > > + mutex_lock(&kvm->lock); > > + r = -EEXIST; > > + if (lapic_in_kernel(kvm)) > > + goto split_irqchip_unlock; > > + r = -EINVAL; > > + if (atomic_read(&kvm->online_vcpus)) > > + goto split_irqchip_unlock; > > + r = kvm_setup_empty_irq_routing(kvm); > > + if (r) > > + goto split_irqchip_unlock; > > + kvm->arch.irqchip_split = true; > > + r = 0; > > +split_irqchip_unlock: > > + mutex_unlock(&kvm->lock); > > + break; > > + } > > + > > default: > > r = kvm_vm_ioctl_assigned_device(kvm, ioctl, arg); > > } > > @@ -5893,7 +5912,7 @@ void kvm_arch_exit(void) > > int kvm_vcpu_halt(struct kvm_vcpu *vcpu) > > { > > ++vcpu->stat.halt_exits; > > - if (irqchip_in_kernel(vcpu->kvm)) { > > + if (lapic_in_kernel(vcpu->kvm)) { > > vcpu->arch.mp_state = KVM_MP_STATE_HALTED; > > return 1; > > } else { > > @@ -6060,7 +6079,7 @@ static int emulator_fix_hypercall(struct x86_emulate_ctxt *ctxt) > > */ > > static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu) > > { > > - return (!irqchip_in_kernel(vcpu->kvm) && !kvm_cpu_has_interrupt(vcpu) && > > + return (!lapic_in_kernel(vcpu->kvm) && !kvm_cpu_has_interrupt(vcpu) && > > vcpu->run->request_interrupt_window && > > kvm_arch_interrupt_allowed(vcpu)); > > } > > @@ -6072,7 +6091,7 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu) > > kvm_run->if_flag = (kvm_get_rflags(vcpu) & X86_EFLAGS_IF) != 0; > > kvm_run->cr8 = kvm_get_cr8(vcpu); > > kvm_run->apic_base = kvm_get_apic_base(vcpu); > > - if (irqchip_in_kernel(vcpu->kvm)) > > + if (lapic_in_kernel(vcpu->kvm)) > > kvm_run->ready_for_interrupt_injection = 1; > > else > > kvm_run->ready_for_interrupt_injection = > > @@ -6219,7 +6238,7 @@ void kvm_vcpu_reload_apic_access_page(struct kvm_vcpu *vcpu) > > { > > struct page *page = NULL; > > > > - if (!irqchip_in_kernel(vcpu->kvm)) > > + if (!lapic_in_kernel(vcpu->kvm)) > > return; > > > > if (!kvm_x86_ops->set_apic_access_page_addr) > > @@ -6255,7 +6274,7 @@ void kvm_arch_mmu_notifier_invalidate_page(struct kvm *kvm, > > static int vcpu_enter_guest(struct kvm_vcpu *vcpu) > > { > > int r; > > - bool req_int_win = !irqchip_in_kernel(vcpu->kvm) && > > + bool req_int_win = !lapic_in_kernel(vcpu->kvm) && > > vcpu->run->request_interrupt_window; > > bool req_immediate_exit = false; > > > > @@ -6644,7 +6663,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) > > } > > > > /* re-sync apic's tpr */ > > - if (!irqchip_in_kernel(vcpu->kvm)) { > > + if (!lapic_in_kernel(vcpu->kvm)) { > > if (kvm_set_cr8(vcpu, kvm_run->cr8) != 0) { > > r = -EINVAL; > > goto out; > > @@ -7340,7 +7359,7 @@ void kvm_arch_check_processor_compat(void *rtn) > > > > bool kvm_vcpu_compatible(struct kvm_vcpu *vcpu) > > { > > - return irqchip_in_kernel(vcpu->kvm) == (vcpu->arch.apic != NULL); > > + return lapic_in_kernel(vcpu->kvm) == (vcpu->arch.apic != NULL); > > } > > > > struct static_key kvm_no_apic_vcpu __read_mostly; > > @@ -7356,7 +7375,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) > > > > vcpu->arch.pv.pv_unhalted = false; > > vcpu->arch.emulate_ctxt.ops = &emulate_ops; > > - if (!irqchip_in_kernel(kvm) || kvm_vcpu_is_reset_bsp(vcpu)) > > + if (!lapic_in_kernel(kvm) || kvm_vcpu_is_reset_bsp(vcpu)) > > vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE; > > else > > vcpu->arch.mp_state = KVM_MP_STATE_UNINITIALIZED; > > @@ -7374,7 +7393,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) > > if (r < 0) > > goto fail_free_pio_data; > > > > - if (irqchip_in_kernel(kvm)) { > > + if (lapic_in_kernel(kvm)) { > > r = kvm_create_lapic(vcpu); > > if (r < 0) > > goto fail_mmu_destroy; > > @@ -7437,7 +7456,7 @@ void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) > > kvm_mmu_destroy(vcpu); > > srcu_read_unlock(&vcpu->kvm->srcu, idx); > > free_page((unsigned long)vcpu->arch.pio_data); > > - if (!irqchip_in_kernel(vcpu->kvm)) > > + if (!lapic_in_kernel(vcpu->kvm)) > > static_key_slow_dec(&kvm_no_apic_vcpu); > > } > > > > diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h > > index 133ea00..ffe1f4e 100644 > > --- a/include/kvm/arm_vgic.h > > +++ b/include/kvm/arm_vgic.h > > @@ -329,6 +329,7 @@ int kvm_vgic_vcpu_pending_irq(struct kvm_vcpu *vcpu); > > int kvm_vgic_vcpu_active_irq(struct kvm_vcpu *vcpu); > > > > #define irqchip_in_kernel(k) (!!((k)->arch.vgic.in_kernel)) > > +#define lapic_in_kernel(k) (irqchip_in_kernel(k)) > > #define vgic_initialized(k) (!!((k)->arch.vgic.nr_cpus)) > > #define vgic_ready(k) ((k)->arch.vgic.ready) > > > > diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h > > index 87fd74a..277b7a1 100644 > > --- a/include/linux/kvm_host.h > > +++ b/include/linux/kvm_host.h > > @@ -934,6 +934,7 @@ static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq) > > #endif > > > > int kvm_setup_default_irq_routing(struct kvm *kvm); > > +int kvm_setup_empty_irq_routing(struct kvm *kvm); > > int kvm_set_irq_routing(struct kvm *kvm, > > const struct kvm_irq_routing_entry *entries, > > unsigned nr, > > diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h > > index 75bd9f7..7d06dc4 100644 > > --- a/include/uapi/linux/kvm.h > > +++ b/include/uapi/linux/kvm.h > > @@ -815,6 +815,7 @@ struct kvm_ppc_smmu_info { > > #define KVM_CAP_S390_IRQ_STATE 114 > > #define KVM_CAP_PPC_HWRNG 115 > > #define KVM_CAP_DISABLE_QUIRKS 116 > > +#define KVM_CAP_SPLIT_IRQCHIP 117 > > > > #ifdef KVM_CAP_IRQ_ROUTING > > > > @@ -1200,6 +1201,8 @@ struct kvm_s390_ucas_mapping { > > /* Available with KVM_CAP_S390_IRQ_STATE */ > > #define KVM_S390_SET_IRQ_STATE _IOW(KVMIO, 0xb5, struct kvm_s390_irq_state) > > #define KVM_S390_GET_IRQ_STATE _IOW(KVMIO, 0xb6, struct kvm_s390_irq_state) > > +/* Available with KVM_CAP_SPLIT_IRQCHIP */ > > +#define KVM_CREATE_SPLIT_IRQCHIP _IO(KVMIO, 0xb7) > > > > #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0) > > #define KVM_DEV_ASSIGN_PCI_2_3 (1 << 1) > > diff --git a/virt/kvm/irqchip.c b/virt/kvm/irqchip.c > > index 1d56a90..8aaceed 100644 > > --- a/virt/kvm/irqchip.c > > +++ b/virt/kvm/irqchip.c > > @@ -73,7 +73,7 @@ int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi) > > { > > struct kvm_kernel_irq_routing_entry route; > > > > - if (!irqchip_in_kernel(kvm) || msi->flags != 0) > > + if (!lapic_in_kernel(kvm) || msi->flags != 0) > > return -EINVAL; > > > > route.msi.address_lo = msi->address_lo; > >