[FYI PATCH 00/20] kvm/queue will be merged soon to kvm/next

[FYI PATCH 00/20] kvm/queue will be merged soon to kvm/next

[Paolo Bonzini]

Hi all,

these are the patches that will be merged Real Soon Now(tm) to kvm/next.
Since all of them have been lying around for a while, I am sending them
out again for everyone's information.

Thanks,

Paolo

Andrey Smetanin (3):
  kvm/x86: Hyper-V HV_X64_MSR_RESET msr
  kvm/x86: Hyper-V HV_X64_MSR_VP_INDEX export for QEMU.
  kvm/x86: Hyper-V HV_X64_MSR_VP_RUNTIME support

Jason Wang (3):
  kvm: use kmalloc() instead of kzalloc() during iodev
    register/unregister
  kvm: add tracepoint for fast mmio
  kvm: add capability for any-length ioeventfds

Paolo Bonzini (6):
  KVM: x86: set TMR when the interrupt is accepted
  KVM: x86: store IOAPIC-handled vectors in each VCPU
  KVM: x86: replace vm_has_apicv hook with cpu_uses_apicv
  KVM: x86: introduce lapic_in_kernel
  KVM: x86: unify handling of interrupt window
  KVM: vmx: disable posted interrupts if no local APIC

Steve Rutherford (4):
  KVM: x86: Split the APIC from the rest of IRQCHIP.
  KVM: x86: Add KVM exit for IOAPIC EOIs
  KVM: x86: Add EOI exit bitmap inference
  KVM: x86: Add support for local interrupt requests from userspace

Wanpeng Li (4):
  KVM: VMX: adjust interface to allocate/free_vpid
  KVM: VMX: introduce __vmx_flush_tlb to handle specific vpid
  KVM: nVMX: emulate the INVVPID instruction
  KVM: nVMX: nested VPID emulation

 Documentation/virtual/kvm/api.txt  |  52 ++++++++++--
 arch/x86/include/asm/kvm_host.h    |  12 ++-
 arch/x86/include/asm/vmx.h         |   1 +
 arch/x86/include/uapi/asm/hyperv.h |   6 ++
 arch/x86/kvm/hyperv.c              |  31 ++++++-
 arch/x86/kvm/i8254.c               |   4 +-
 arch/x86/kvm/ioapic.c              |  27 +-----
 arch/x86/kvm/ioapic.h              |  15 ++--
 arch/x86/kvm/irq.c                 |  40 ++++++---
 arch/x86/kvm/irq.h                 |  27 +++++-
 arch/x86/kvm/irq_comm.c            |  51 ++++++++++-
 arch/x86/kvm/lapic.c               |  64 +++++++++-----
 arch/x86/kvm/lapic.h               |   5 +-
 arch/x86/kvm/mmu.c                 |   2 +-
 arch/x86/kvm/svm.c                 |  21 +----
 arch/x86/kvm/trace.h               |  18 ++++
 arch/x86/kvm/vmx.c                 | 168 +++++++++++++++++++++++--------------
 arch/x86/kvm/x86.c                 | 126 +++++++++++++++++++++-------
 include/linux/kvm_host.h           |  21 ++++-
 include/uapi/linux/kvm.h           |   7 ++
 kernel/sched/cputime.c             |   2 +
 virt/kvm/eventfd.c                 |   4 +-
 virt/kvm/irqchip.c                 |  12 +--
 virt/kvm/kvm_main.c                |   5 +-
 24 files changed, 516 insertions(+), 205 deletions(-)

-- 
1.8.3.1

[FYI PATCH 01/20] KVM: x86: set TMR when the interrupt is accepted

[Paolo Bonzini]

Do not compute TMR in advance.  Instead, set the TMR just before the interrupt
is accepted into the IRR.  This limits the coupling between IOAPIC and LAPIC.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/ioapic.c |  9 ++-------
 arch/x86/kvm/ioapic.h |  3 +--
 arch/x86/kvm/lapic.c  | 19 ++++++++++---------
 arch/x86/kvm/lapic.h  |  1 -
 arch/x86/kvm/x86.c    |  5 +----
 5 files changed, 14 insertions(+), 23 deletions(-)

diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c
index 856f79105bb5..eaf4ec38d980 100644
--- a/arch/x86/kvm/ioapic.c
+++ b/arch/x86/kvm/ioapic.c
@@ -246,8 +246,7 @@ static void update_handled_vectors(struct kvm_ioapic *ioapic)
 	smp_wmb();
 }
 
-void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap,
-			u32 *tmr)
+void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
 {
 	struct kvm_ioapic *ioapic = vcpu->kvm->arch.vioapic;
 	union kvm_ioapic_redirect_entry *e;
@@ -260,13 +259,9 @@ void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap,
 		    kvm_irq_has_notifier(ioapic->kvm, KVM_IRQCHIP_IOAPIC, index) ||
 		    index == RTC_GSI) {
 			if (kvm_apic_match_dest(vcpu, NULL, 0,
-				e->fields.dest_id, e->fields.dest_mode)) {
+				e->fields.dest_id, e->fields.dest_mode))
 				__set_bit(e->fields.vector,
 					(unsigned long *)eoi_exit_bitmap);
-				if (e->fields.trig_mode == IOAPIC_LEVEL_TRIG)
-					__set_bit(e->fields.vector,
-						(unsigned long *)tmr);
-			}
 		}
 	}
 	spin_unlock(&ioapic->lock);
diff --git a/arch/x86/kvm/ioapic.h b/arch/x86/kvm/ioapic.h
index ca0b0b4e6256..3dbd0e2aac4e 100644
--- a/arch/x86/kvm/ioapic.h
+++ b/arch/x86/kvm/ioapic.h
@@ -120,7 +120,6 @@ int kvm_irq_delivery_to_apic(struct kvm *kvm, struct kvm_lapic *src,
 		struct kvm_lapic_irq *irq, unsigned long *dest_map);
 int kvm_get_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
 int kvm_set_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
-void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap,
-			u32 *tmr);
+void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap);
 
 #endif
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 8d9013c5e1ee..5693dd9fc163 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -551,15 +551,6 @@ static void pv_eoi_clr_pending(struct kvm_vcpu *vcpu)
 	__clear_bit(KVM_APIC_PV_EOI_PENDING, &vcpu->arch.apic_attention);
 }
 
-void kvm_apic_update_tmr(struct kvm_vcpu *vcpu, u32 *tmr)
-{
-	struct kvm_lapic *apic = vcpu->arch.apic;
-	int i;
-
-	for (i = 0; i < 8; i++)
-		apic_set_reg(apic, APIC_TMR + 0x10 * i, tmr[i]);
-}
-
 static void apic_update_ppr(struct kvm_lapic *apic)
 {
 	u32 tpr, isrv, ppr, old_ppr;
@@ -781,6 +772,9 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 	case APIC_DM_LOWEST:
 		vcpu->arch.apic_arb_prio++;
 	case APIC_DM_FIXED:
+		if (unlikely(trig_mode && !level))
+			break;
+
 		/* FIXME add logic for vcpu on reset */
 		if (unlikely(!apic_enabled(apic)))
 			break;
@@ -790,6 +784,13 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 		if (dest_map)
 			__set_bit(vcpu->vcpu_id, dest_map);
 
+		if (apic_test_vector(vector, apic->regs + APIC_TMR) != !!trig_mode) {
+			if (trig_mode)
+				apic_set_vector(vector, apic->regs + APIC_TMR);
+			else
+				apic_clear_vector(vector, apic->regs + APIC_TMR);
+		}
+
 		if (kvm_x86_ops->deliver_posted_interrupt)
 			kvm_x86_ops->deliver_posted_interrupt(vcpu, vector);
 		else {
diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h
index 764037991d26..eb46d6bcaa75 100644
--- a/arch/x86/kvm/lapic.h
+++ b/arch/x86/kvm/lapic.h
@@ -57,7 +57,6 @@ void kvm_lapic_set_base(struct kvm_vcpu *vcpu, u64 value);
 u64 kvm_lapic_get_base(struct kvm_vcpu *vcpu);
 void kvm_apic_set_version(struct kvm_vcpu *vcpu);
 
-void kvm_apic_update_tmr(struct kvm_vcpu *vcpu, u32 *tmr);
 void __kvm_apic_update_irr(u32 *pir, void *regs);
 void kvm_apic_update_irr(struct kvm_vcpu *vcpu, u32 *pir);
 int kvm_apic_set_irq(struct kvm_vcpu *vcpu, struct kvm_lapic_irq *irq,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 991466bf8dee..c0ffc636fc8b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6148,17 +6148,14 @@ static void process_smi(struct kvm_vcpu *vcpu)
 static void vcpu_scan_ioapic(struct kvm_vcpu *vcpu)
 {
 	u64 eoi_exit_bitmap[4];
-	u32 tmr[8];
 
 	if (!kvm_apic_hw_enabled(vcpu->arch.apic))
 		return;
 
 	memset(eoi_exit_bitmap, 0, 32);
-	memset(tmr, 0, 32);
 
-	kvm_ioapic_scan_entry(vcpu, eoi_exit_bitmap, tmr);
+	kvm_ioapic_scan_entry(vcpu, eoi_exit_bitmap);
 	kvm_x86_ops->load_eoi_exitmap(vcpu, eoi_exit_bitmap);
-	kvm_apic_update_tmr(vcpu, tmr);
 }
 
 static void kvm_vcpu_flush_tlb(struct kvm_vcpu *vcpu)
-- 
1.8.3.1

[FYI PATCH 02/20] KVM: x86: store IOAPIC-handled vectors in each VCPU

[Paolo Bonzini]

We can reuse the algorithm that computes the EOI exit bitmap to figure
out which vectors are handled by the IOAPIC.  The only difference
between the two is for edge-triggered interrupts other than IRQ8
that have no notifiers active; however, the IOAPIC does not have to
do anything special for these interrupts anyway.

This again limits the interactions between the IOAPIC and the LAPIC,
making it easier to move the former to userspace.

Inspired by a patch from Steve Rutherford.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/asm/kvm_host.h |  3 ++-
 arch/x86/kvm/ioapic.c           | 18 ++----------------
 arch/x86/kvm/ioapic.h           |  8 --------
 arch/x86/kvm/lapic.c            | 10 ++++++++--
 arch/x86/kvm/svm.c              |  2 +-
 arch/x86/kvm/vmx.c              |  3 ++-
 arch/x86/kvm/x86.c              |  8 +++-----
 7 files changed, 18 insertions(+), 34 deletions(-)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 2beee0382088..33609c2c743b 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -396,6 +396,7 @@ struct kvm_vcpu_arch {
 	u64 efer;
 	u64 apic_base;
 	struct kvm_lapic *apic;    /* kernel irqchip context */
+	u64 eoi_exit_bitmap[4];
 	unsigned long apic_attention;
 	int32_t apic_arb_prio;
 	int mp_state;
@@ -822,7 +823,7 @@ struct kvm_x86_ops {
 	int (*vm_has_apicv)(struct kvm *kvm);
 	void (*hwapic_irr_update)(struct kvm_vcpu *vcpu, int max_irr);
 	void (*hwapic_isr_update)(struct kvm *kvm, int isr);
-	void (*load_eoi_exitmap)(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap);
+	void (*load_eoi_exitmap)(struct kvm_vcpu *vcpu);
 	void (*set_virtual_x2apic_mode)(struct kvm_vcpu *vcpu, bool set);
 	void (*set_apic_access_page_addr)(struct kvm_vcpu *vcpu, hpa_t hpa);
 	void (*deliver_posted_interrupt)(struct kvm_vcpu *vcpu, int vector);
diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c
index eaf4ec38d980..2dcda0f188ba 100644
--- a/arch/x86/kvm/ioapic.c
+++ b/arch/x86/kvm/ioapic.c
@@ -233,19 +233,6 @@ static void kvm_ioapic_inject_all(struct kvm_ioapic *ioapic, unsigned long irr)
 }
 
 
-static void update_handled_vectors(struct kvm_ioapic *ioapic)
-{
-	DECLARE_BITMAP(handled_vectors, 256);
-	int i;
-
-	memset(handled_vectors, 0, sizeof(handled_vectors));
-	for (i = 0; i < IOAPIC_NUM_PINS; ++i)
-		__set_bit(ioapic->redirtbl[i].fields.vector, handled_vectors);
-	memcpy(ioapic->handled_vectors, handled_vectors,
-	       sizeof(handled_vectors));
-	smp_wmb();
-}
-
 void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
 {
 	struct kvm_ioapic *ioapic = vcpu->kvm->arch.vioapic;
@@ -310,7 +297,6 @@ static void ioapic_write_indirect(struct kvm_ioapic *ioapic, u32 val)
 			e->bits |= (u32) val;
 			e->fields.remote_irr = 0;
 		}
-		update_handled_vectors(ioapic);
 		mask_after = e->fields.mask;
 		if (mask_before != mask_after)
 			kvm_fire_mask_notifiers(ioapic->kvm, KVM_IRQCHIP_IOAPIC, index, mask_after);
@@ -594,7 +580,6 @@ static void kvm_ioapic_reset(struct kvm_ioapic *ioapic)
 	ioapic->id = 0;
 	memset(ioapic->irq_eoi, 0x00, IOAPIC_NUM_PINS);
 	rtc_irq_eoi_tracking_reset(ioapic);
-	update_handled_vectors(ioapic);
 }
 
 static const struct kvm_io_device_ops ioapic_mmio_ops = {
@@ -623,8 +608,10 @@ int kvm_ioapic_init(struct kvm *kvm)
 	if (ret < 0) {
 		kvm->arch.vioapic = NULL;
 		kfree(ioapic);
+		return ret;
 	}
 
+	kvm_vcpu_request_scan_ioapic(kvm);
 	return ret;
 }
 
@@ -661,7 +648,6 @@ int kvm_set_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state)
 	memcpy(ioapic, state, sizeof(struct kvm_ioapic_state));
 	ioapic->irr = 0;
 	ioapic->irr_delivered = 0;
-	update_handled_vectors(ioapic);
 	kvm_vcpu_request_scan_ioapic(kvm);
 	kvm_ioapic_inject_all(ioapic, state->irr);
 	spin_unlock(&ioapic->lock);
diff --git a/arch/x86/kvm/ioapic.h b/arch/x86/kvm/ioapic.h
index 3dbd0e2aac4e..bf36d66a1951 100644
--- a/arch/x86/kvm/ioapic.h
+++ b/arch/x86/kvm/ioapic.h
@@ -73,7 +73,6 @@ struct kvm_ioapic {
 	struct kvm *kvm;
 	void (*ack_notifier)(void *opaque, int irq);
 	spinlock_t lock;
-	DECLARE_BITMAP(handled_vectors, 256);
 	struct rtc_status rtc_status;
 	struct delayed_work eoi_inject;
 	u32 irq_eoi[IOAPIC_NUM_PINS];
@@ -98,13 +97,6 @@ static inline struct kvm_ioapic *ioapic_irqchip(struct kvm *kvm)
 	return kvm->arch.vioapic;
 }
 
-static inline bool kvm_ioapic_handles_vector(struct kvm *kvm, int vector)
-{
-	struct kvm_ioapic *ioapic = kvm->arch.vioapic;
-	smp_rmb();
-	return test_bit(vector, ioapic->handled_vectors);
-}
-
 void kvm_rtc_eoi_tracking_restore_one(struct kvm_vcpu *vcpu);
 bool kvm_apic_match_dest(struct kvm_vcpu *vcpu, struct kvm_lapic *source,
 		int short_hand, unsigned int dest, int dest_mode);
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 5693dd9fc163..4c30fb0a48a1 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -869,14 +869,20 @@ int kvm_apic_compare_prio(struct kvm_vcpu *vcpu1, struct kvm_vcpu *vcpu2)
 	return vcpu1->arch.apic_arb_prio - vcpu2->arch.apic_arb_prio;
 }
 
+static bool kvm_ioapic_handles_vector(struct kvm_lapic *apic, int vector)
+{
+	return test_bit(vector, (ulong *)apic->vcpu->arch.eoi_exit_bitmap);
+}
+
 static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector)
 {
-	if (kvm_ioapic_handles_vector(apic->vcpu->kvm, vector)) {
+	if (kvm_ioapic_handles_vector(apic, vector)) {
 		int trigger_mode;
 		if (apic_test_vector(vector, apic->regs + APIC_TMR))
 			trigger_mode = IOAPIC_LEVEL_TRIG;
 		else
 			trigger_mode = IOAPIC_EDGE_TRIG;
+
 		kvm_ioapic_update_eoi(apic->vcpu, vector, trigger_mode);
 	}
 }
@@ -1923,7 +1929,7 @@ 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(apic, apic->highest_isr_cache)) {
 		/*
 		 * PV EOI was disabled by apic_sync_pv_eoi_from_guest
 		 * so we need not do anything here.
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 94b7d15db3fc..79612964e8f1 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3768,7 +3768,7 @@ static int svm_vm_has_apicv(struct kvm *kvm)
 	return 0;
 }
 
-static void svm_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
+static void svm_load_eoi_exitmap(struct kvm_vcpu *vcpu)
 {
 	return;
 }
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 64076740251e..82ea70a8c9e7 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -8043,8 +8043,9 @@ static void vmx_hwapic_irr_update(struct kvm_vcpu *vcpu, int max_irr)
 	}
 }
 
-static void vmx_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
+static void vmx_load_eoi_exitmap(struct kvm_vcpu *vcpu)
 {
+	u64 *eoi_exit_bitmap = vcpu->arch.eoi_exit_bitmap;
 	if (!vmx_vm_has_apicv(vcpu->kvm))
 		return;
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c0ffc636fc8b..7a9af61dad07 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6147,15 +6147,13 @@ static void process_smi(struct kvm_vcpu *vcpu)
 
 static void vcpu_scan_ioapic(struct kvm_vcpu *vcpu)
 {
-	u64 eoi_exit_bitmap[4];
-
 	if (!kvm_apic_hw_enabled(vcpu->arch.apic))
 		return;
 
-	memset(eoi_exit_bitmap, 0, 32);
+	memset(vcpu->arch.eoi_exit_bitmap, 0, 256 / 8);
 
-	kvm_ioapic_scan_entry(vcpu, eoi_exit_bitmap);
-	kvm_x86_ops->load_eoi_exitmap(vcpu, eoi_exit_bitmap);
+	kvm_ioapic_scan_entry(vcpu, vcpu->arch.eoi_exit_bitmap);
+	kvm_x86_ops->load_eoi_exitmap(vcpu);
 }
 
 static void kvm_vcpu_flush_tlb(struct kvm_vcpu *vcpu)
-- 
1.8.3.1

[FYI PATCH 03/20] KVM: x86: replace vm_has_apicv hook with cpu_uses_apicv

[Paolo Bonzini]

This will avoid an unnecessary trip to ->kvm and from there to the VPIC.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/asm/kvm_host.h | 2 +-
 arch/x86/kvm/irq.c              | 2 +-
 arch/x86/kvm/lapic.c            | 4 ++--
 arch/x86/kvm/lapic.h            | 4 ++--
 arch/x86/kvm/svm.c              | 4 ++--
 arch/x86/kvm/vmx.c              | 8 +++++++-
 6 files changed, 15 insertions(+), 9 deletions(-)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 33609c2c743b..a0ef289d5a86 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -820,7 +820,7 @@ 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);
-	int (*vm_has_apicv)(struct kvm *kvm);
+	int (*cpu_uses_apicv)(struct kvm_vcpu *vcpu);
 	void (*hwapic_irr_update)(struct kvm_vcpu *vcpu, int max_irr);
 	void (*hwapic_isr_update)(struct kvm *kvm, int isr);
 	void (*load_eoi_exitmap)(struct kvm_vcpu *vcpu);
diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c
index a1ec6a50a05a..c0dad893dc59 100644
--- a/arch/x86/kvm/irq.c
+++ b/arch/x86/kvm/irq.c
@@ -63,7 +63,7 @@ int kvm_cpu_has_injectable_intr(struct kvm_vcpu *v)
 	if (kvm_cpu_has_extint(v))
 		return 1;
 
-	if (kvm_apic_vid_enabled(v->kvm))
+	if (kvm_vcpu_apic_vid_enabled(v))
 		return 0;
 
 	return kvm_apic_has_interrupt(v) != -1; /* LAPIC */
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 4c30fb0a48a1..c568d69c7060 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -390,7 +390,7 @@ static inline void apic_clear_irr(int vec, struct kvm_lapic *apic)
 
 	vcpu = apic->vcpu;
 
-	if (unlikely(kvm_apic_vid_enabled(vcpu->kvm))) {
+	if (unlikely(kvm_vcpu_apic_vid_enabled(vcpu))) {
 		/* try to update RVI */
 		apic_clear_vector(vec, apic->regs + APIC_IRR);
 		kvm_make_request(KVM_REQ_EVENT, vcpu);
@@ -1622,7 +1622,7 @@ void kvm_lapic_reset(struct kvm_vcpu *vcpu, bool init_event)
 		apic_set_reg(apic, APIC_ISR + 0x10 * i, 0);
 		apic_set_reg(apic, APIC_TMR + 0x10 * i, 0);
 	}
-	apic->irr_pending = kvm_apic_vid_enabled(vcpu->kvm);
+	apic->irr_pending = kvm_vcpu_apic_vid_enabled(vcpu);
 	apic->isr_count = kvm_x86_ops->hwapic_isr_update ? 1 : 0;
 	apic->highest_isr_cache = -1;
 	update_divide_count(apic);
diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h
index eb46d6bcaa75..7259d272416f 100644
--- a/arch/x86/kvm/lapic.h
+++ b/arch/x86/kvm/lapic.h
@@ -143,9 +143,9 @@ static inline int apic_x2apic_mode(struct kvm_lapic *apic)
 	return apic->vcpu->arch.apic_base & X2APIC_ENABLE;
 }
 
-static inline bool kvm_apic_vid_enabled(struct kvm *kvm)
+static inline bool kvm_vcpu_apic_vid_enabled(struct kvm_vcpu *vcpu)
 {
-	return kvm_x86_ops->vm_has_apicv(kvm);
+	return kvm_x86_ops->cpu_uses_apicv(vcpu);
 }
 
 static inline bool kvm_apic_has_events(struct kvm_vcpu *vcpu)
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 79612964e8f1..98889c882ced 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3763,7 +3763,7 @@ static void svm_set_virtual_x2apic_mode(struct kvm_vcpu *vcpu, bool set)
 	return;
 }
 
-static int svm_vm_has_apicv(struct kvm *kvm)
+static int svm_cpu_uses_apicv(struct kvm_vcpu *vcpu)
 {
 	return 0;
 }
@@ -4524,7 +4524,7 @@ static struct kvm_x86_ops svm_x86_ops = {
 	.enable_irq_window = enable_irq_window,
 	.update_cr8_intercept = update_cr8_intercept,
 	.set_virtual_x2apic_mode = svm_set_virtual_x2apic_mode,
-	.vm_has_apicv = svm_vm_has_apicv,
+	.cpu_uses_apicv = svm_cpu_uses_apicv,
 	.load_eoi_exitmap = svm_load_eoi_exitmap,
 	.sync_pir_to_irr = svm_sync_pir_to_irr,
 
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 82ea70a8c9e7..32a38494dd6f 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -810,6 +810,7 @@ static void kvm_cpu_vmxoff(void);
 static bool vmx_mpx_supported(void);
 static bool vmx_xsaves_supported(void);
 static int vmx_vm_has_apicv(struct kvm *kvm);
+static int vmx_cpu_uses_apicv(struct kvm_vcpu *vcpu);
 static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr);
 static void vmx_set_segment(struct kvm_vcpu *vcpu,
 			    struct kvm_segment *var, int seg);
@@ -4337,6 +4338,11 @@ static int vmx_vm_has_apicv(struct kvm *kvm)
 	return enable_apicv && irqchip_in_kernel(kvm);
 }
 
+static int vmx_cpu_uses_apicv(struct kvm_vcpu *vcpu)
+{
+	return vmx_vm_has_apicv(vcpu->kvm);
+}
+
 static int vmx_complete_nested_posted_interrupt(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
@@ -10357,7 +10363,7 @@ static struct kvm_x86_ops vmx_x86_ops = {
 	.update_cr8_intercept = update_cr8_intercept,
 	.set_virtual_x2apic_mode = vmx_set_virtual_x2apic_mode,
 	.set_apic_access_page_addr = vmx_set_apic_access_page_addr,
-	.vm_has_apicv = vmx_vm_has_apicv,
+	.cpu_uses_apicv = vmx_cpu_uses_apicv,
 	.load_eoi_exitmap = vmx_load_eoi_exitmap,
 	.hwapic_irr_update = vmx_hwapic_irr_update,
 	.hwapic_isr_update = vmx_hwapic_isr_update,
-- 
1.8.3.1

[FYI PATCH 04/20] KVM: x86: introduce lapic_in_kernel

[Paolo Bonzini]

Avoid pointer chasing and memory barriers, and simplify the code
when split irqchip (LAPIC in kernel, IOAPIC/PIC in userspace)
is introduced.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/irq.c   |  6 +++---
 arch/x86/kvm/irq.h   |  8 ++++++++
 arch/x86/kvm/lapic.c |  4 ++--
 arch/x86/kvm/mmu.c   |  2 +-
 arch/x86/kvm/svm.c   |  4 ++--
 arch/x86/kvm/vmx.c   | 46 ++++++++++++++++++++--------------------------
 arch/x86/kvm/x86.c   | 18 +++++++++---------
 7 files changed, 45 insertions(+), 43 deletions(-)

diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c
index c0dad893dc59..b653ae202c8e 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))
 		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))
 		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))
 		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 3d782a2c336a..9e6e7e04de98 100644
--- a/arch/x86/kvm/irq.h
+++ b/arch/x86/kvm/irq.h
@@ -92,6 +92,14 @@ static inline int irqchip_in_kernel(struct kvm *kvm)
 	return vpic != NULL;
 }
 
+static inline int lapic_in_kernel(struct kvm_vcpu *vcpu)
+{
+	/* Same as irqchip_in_kernel(vcpu->kvm), but with less
+	 * pointer chasing and no unnecessary memory barriers.
+	 */
+	return vcpu->arch.apic != NULL;
+}
+
 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/lapic.c b/arch/x86/kvm/lapic.c
index c568d69c7060..c4bcc86d6dc4 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1985,7 +1985,7 @@ 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) || !apic_x2apic_mode(apic))
 		return 1;
 
 	if (reg == APIC_ICR2)
@@ -2002,7 +2002,7 @@ 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) || !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 ff606f507913..c3f39aa9b9cb 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -3427,7 +3427,7 @@ 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_event_needs_reinjection(vcpu)))
 		return false;
 
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 98889c882ced..89d278a0ad37 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3154,7 +3154,7 @@ 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))
 		return r;
 	if (cr8_prev <= kvm_get_cr8(&svm->vcpu))
 		return r;
@@ -3409,7 +3409,7 @@ 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_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 32a38494dd6f..d5b87be89631 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -809,7 +809,6 @@ static void kvm_cpu_vmxon(u64 addr);
 static void kvm_cpu_vmxoff(void);
 static bool vmx_mpx_supported(void);
 static bool vmx_xsaves_supported(void);
-static int vmx_vm_has_apicv(struct kvm *kvm);
 static int vmx_cpu_uses_apicv(struct kvm_vcpu *vcpu);
 static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr);
 static void vmx_set_segment(struct kvm_vcpu *vcpu,
@@ -947,9 +946,9 @@ static inline bool cpu_has_vmx_tpr_shadow(void)
 	return vmcs_config.cpu_based_exec_ctrl & CPU_BASED_TPR_SHADOW;
 }
 
-static inline bool vm_need_tpr_shadow(struct kvm *kvm)
+static inline bool cpu_need_tpr_shadow(struct kvm_vcpu *vcpu)
 {
-	return (cpu_has_vmx_tpr_shadow()) && (irqchip_in_kernel(kvm));
+	return cpu_has_vmx_tpr_shadow() && lapic_in_kernel(vcpu);
 }
 
 static inline bool cpu_has_secondary_exec_ctrls(void)
@@ -1063,9 +1062,9 @@ static inline bool cpu_has_vmx_ple(void)
 		SECONDARY_EXEC_PAUSE_LOOP_EXITING;
 }
 
-static inline bool vm_need_virtualize_apic_accesses(struct kvm *kvm)
+static inline bool cpu_need_virtualize_apic_accesses(struct kvm_vcpu *vcpu)
 {
-	return flexpriority_enabled && irqchip_in_kernel(kvm);
+	return flexpriority_enabled && lapic_in_kernel(vcpu);
 }
 
 static inline bool cpu_has_vmx_vpid(void)
@@ -2378,7 +2377,7 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx)
 	vmx->nested.nested_vmx_pinbased_ctls_high |=
 		PIN_BASED_ALWAYSON_WITHOUT_TRUE_MSR |
 		PIN_BASED_VMX_PREEMPTION_TIMER;
-	if (vmx_vm_has_apicv(vmx->vcpu.kvm))
+	if (vmx_cpu_uses_apicv(&vmx->vcpu))
 		vmx->nested.nested_vmx_pinbased_ctls_high |=
 			PIN_BASED_POSTED_INTR;
 
@@ -4333,14 +4332,9 @@ static void vmx_disable_intercept_msr_write_x2apic(u32 msr)
 			msr, MSR_TYPE_W);
 }
 
-static int vmx_vm_has_apicv(struct kvm *kvm)
-{
-	return enable_apicv && irqchip_in_kernel(kvm);
-}
-
 static int vmx_cpu_uses_apicv(struct kvm_vcpu *vcpu)
 {
-	return vmx_vm_has_apicv(vcpu->kvm);
+	return enable_apicv && lapic_in_kernel(vcpu);
 }
 
 static int vmx_complete_nested_posted_interrupt(struct kvm_vcpu *vcpu)
@@ -4520,7 +4514,7 @@ static u32 vmx_pin_based_exec_ctrl(struct vcpu_vmx *vmx)
 {
 	u32 pin_based_exec_ctrl = vmcs_config.pin_based_exec_ctrl;
 
-	if (!vmx_vm_has_apicv(vmx->vcpu.kvm))
+	if (!vmx_cpu_uses_apicv(&vmx->vcpu))
 		pin_based_exec_ctrl &= ~PIN_BASED_POSTED_INTR;
 	return pin_based_exec_ctrl;
 }
@@ -4532,7 +4526,7 @@ static u32 vmx_exec_control(struct vcpu_vmx *vmx)
 	if (vmx->vcpu.arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)
 		exec_control &= ~CPU_BASED_MOV_DR_EXITING;
 
-	if (!vm_need_tpr_shadow(vmx->vcpu.kvm)) {
+	if (!cpu_need_tpr_shadow(&vmx->vcpu)) {
 		exec_control &= ~CPU_BASED_TPR_SHADOW;
 #ifdef CONFIG_X86_64
 		exec_control |= CPU_BASED_CR8_STORE_EXITING |
@@ -4549,7 +4543,7 @@ static u32 vmx_exec_control(struct vcpu_vmx *vmx)
 static u32 vmx_secondary_exec_control(struct vcpu_vmx *vmx)
 {
 	u32 exec_control = vmcs_config.cpu_based_2nd_exec_ctrl;
-	if (!vm_need_virtualize_apic_accesses(vmx->vcpu.kvm))
+	if (!cpu_need_virtualize_apic_accesses(&vmx->vcpu))
 		exec_control &= ~SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES;
 	if (vmx->vpid == 0)
 		exec_control &= ~SECONDARY_EXEC_ENABLE_VPID;
@@ -4563,7 +4557,7 @@ static u32 vmx_secondary_exec_control(struct vcpu_vmx *vmx)
 		exec_control &= ~SECONDARY_EXEC_UNRESTRICTED_GUEST;
 	if (!ple_gap)
 		exec_control &= ~SECONDARY_EXEC_PAUSE_LOOP_EXITING;
-	if (!vmx_vm_has_apicv(vmx->vcpu.kvm))
+	if (!vmx_cpu_uses_apicv(&vmx->vcpu))
 		exec_control &= ~(SECONDARY_EXEC_APIC_REGISTER_VIRT |
 				  SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY);
 	exec_control &= ~SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE;
@@ -4624,7 +4618,7 @@ static int vmx_vcpu_setup(struct vcpu_vmx *vmx)
 				vmx_secondary_exec_control(vmx));
 	}
 
-	if (vmx_vm_has_apicv(vmx->vcpu.kvm)) {
+	if (vmx_cpu_uses_apicv(&vmx->vcpu)) {
 		vmcs_write64(EOI_EXIT_BITMAP0, 0);
 		vmcs_write64(EOI_EXIT_BITMAP1, 0);
 		vmcs_write64(EOI_EXIT_BITMAP2, 0);
@@ -4768,7 +4762,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
 
 	if (cpu_has_vmx_tpr_shadow() && !init_event) {
 		vmcs_write64(VIRTUAL_APIC_PAGE_ADDR, 0);
-		if (vm_need_tpr_shadow(vcpu->kvm))
+		if (cpu_need_tpr_shadow(vcpu))
 			vmcs_write64(VIRTUAL_APIC_PAGE_ADDR,
 				     __pa(vcpu->arch.apic->regs));
 		vmcs_write32(TPR_THRESHOLD, 0);
@@ -4776,7 +4770,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
 
 	kvm_make_request(KVM_REQ_APIC_PAGE_RELOAD, vcpu);
 
-	if (vmx_vm_has_apicv(vcpu->kvm))
+	if (vmx_cpu_uses_apicv(vcpu))
 		memset(&vmx->pi_desc, 0, sizeof(struct pi_desc));
 
 	if (vmx->vpid != 0)
@@ -5316,7 +5310,7 @@ 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))
 					return 1;
 				if (cr8_prev <= cr8)
 					return 1;
@@ -5535,7 +5529,7 @@ 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) &&
 	    vcpu->run->request_interrupt_window &&
 	    !kvm_cpu_has_interrupt(vcpu)) {
 		vcpu->run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
@@ -7944,10 +7938,10 @@ static void vmx_set_virtual_x2apic_mode(struct kvm_vcpu *vcpu, bool set)
 	 * apicv
 	 */
 	if (!cpu_has_vmx_virtualize_x2apic_mode() ||
-				!vmx_vm_has_apicv(vcpu->kvm))
+				!vmx_cpu_uses_apicv(vcpu))
 		return;
 
-	if (!vm_need_tpr_shadow(vcpu->kvm))
+	if (!cpu_need_tpr_shadow(vcpu))
 		return;
 
 	sec_exec_control = vmcs_read32(SECONDARY_VM_EXEC_CONTROL);
@@ -8052,7 +8046,7 @@ static void vmx_hwapic_irr_update(struct kvm_vcpu *vcpu, int max_irr)
 static void vmx_load_eoi_exitmap(struct kvm_vcpu *vcpu)
 {
 	u64 *eoi_exit_bitmap = vcpu->arch.eoi_exit_bitmap;
-	if (!vmx_vm_has_apicv(vcpu->kvm))
+	if (!vmx_cpu_uses_apicv(vcpu))
 		return;
 
 	vmcs_write64(EOI_EXIT_BITMAP0, eoi_exit_bitmap[0]);
@@ -8551,7 +8545,7 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 	put_cpu();
 	if (err)
 		goto free_vmcs;
-	if (vm_need_virtualize_apic_accesses(kvm)) {
+	if (cpu_need_virtualize_apic_accesses(&vmx->vcpu)) {
 		err = alloc_apic_access_page(kvm);
 		if (err)
 			goto free_vmcs;
@@ -9339,7 +9333,7 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
 				vmcs_write64(APIC_ACCESS_ADDR,
 				  page_to_phys(vmx->nested.apic_access_page));
 		} else if (!(nested_cpu_has_virt_x2apic_mode(vmcs12)) &&
-			    (vm_need_virtualize_apic_accesses(vmx->vcpu.kvm))) {
+			    cpu_need_virtualize_apic_accesses(&vmx->vcpu)) {
 			exec_control |=
 				SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES;
 			kvm_vcpu_reload_apic_access_page(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 7a9af61dad07..e25bc4e2b7a9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -788,7 +788,7 @@ 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_lapic_set_tpr(vcpu, cr8);
 	else
 		vcpu->arch.cr8 = cr8;
@@ -798,7 +798,7 @@ 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))
 		return kvm_lapic_get_cr8(vcpu);
 	else
 		return vcpu->arch.cr8;
@@ -3179,7 +3179,7 @@ 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))
 			goto out;
 		r = -EFAULT;
 		if (copy_from_user(&va, argp, sizeof va))
@@ -5670,7 +5670,7 @@ 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)) {
 		vcpu->arch.mp_state = KVM_MP_STATE_HALTED;
 		return 1;
 	} else {
@@ -6166,7 +6166,7 @@ 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))
 		return;
 
 	if (!kvm_x86_ops->set_apic_access_page_addr)
@@ -6204,7 +6204,7 @@ 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) &&
 		vcpu->run->request_interrupt_window;
 	bool req_immediate_exit = false;
 
@@ -6601,7 +6601,7 @@ 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)) {
 		if (kvm_set_cr8(vcpu, kvm_run->cr8) != 0) {
 			r = -EINVAL;
 			goto out;
@@ -7301,7 +7301,7 @@ bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu)
 
 bool kvm_vcpu_compatible(struct kvm_vcpu *vcpu)
 {
-	return irqchip_in_kernel(vcpu->kvm) == (vcpu->arch.apic != NULL);
+	return irqchip_in_kernel(vcpu->kvm) == lapic_in_kernel(vcpu);
 }
 
 struct static_key kvm_no_apic_vcpu __read_mostly;
@@ -7395,7 +7395,7 @@ 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))
 		static_key_slow_dec(&kvm_no_apic_vcpu);
 }
 
-- 
1.8.3.1

[FYI PATCH 05/20] KVM: x86: unify handling of interrupt window

[Paolo Bonzini]

The interrupt window is currently checked twice, once in vmx.c/svm.c and
once in dm_request_for_irq_injection.  The only difference is the extra
check for kvm_arch_interrupt_allowed in dm_request_for_irq_injection,
and the different return value (EINTR/KVM_EXIT_INTR for vmx.c/svm.c vs.
0/KVM_EXIT_IRQ_WINDOW_OPEN for dm_request_for_irq_injection).

However, dm_request_for_irq_injection is basically dead code!  Revive it
by removing the checks in vmx.c and svm.c's vmexit handlers, and
fixing the returned values for the dm_request_for_irq_injection case.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/svm.c | 13 -------------
 arch/x86/kvm/vmx.c | 11 -----------
 arch/x86/kvm/x86.c |  4 ++--
 3 files changed, 2 insertions(+), 26 deletions(-)

diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 89d278a0ad37..7d7bbe2651d2 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3398,24 +3398,11 @@ static int msr_interception(struct vcpu_svm *svm)
 
 static int interrupt_window_interception(struct vcpu_svm *svm)
 {
-	struct kvm_run *kvm_run = svm->vcpu.run;
-
 	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
 	svm_clear_vintr(svm);
 	svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
 	mark_dirty(svm->vmcb, VMCB_INTR);
 	++svm->vcpu.stat.irq_window_exits;
-	/*
-	 * If the user space waits to inject interrupts, exit as soon as
-	 * possible
-	 */
-	if (!lapic_in_kernel(&svm->vcpu) &&
-	    kvm_run->request_interrupt_window &&
-	    !kvm_cpu_has_interrupt(&svm->vcpu)) {
-		kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
-		return 0;
-	}
-
 	return 1;
 }
 
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index d5b87be89631..90605f70a7d0 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -5524,17 +5524,6 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu)
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
 	++vcpu->stat.irq_window_exits;
-
-	/*
-	 * If the user space waits to inject interrupts, exit as soon as
-	 * possible
-	 */
-	if (!lapic_in_kernel(vcpu) &&
-	    vcpu->run->request_interrupt_window &&
-	    !kvm_cpu_has_interrupt(vcpu)) {
-		vcpu->run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
-		return 0;
-	}
 	return 1;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index e25bc4e2b7a9..080aaa7a4d91 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6473,8 +6473,8 @@ static int vcpu_run(struct kvm_vcpu *vcpu)
 			kvm_inject_pending_timer_irqs(vcpu);
 
 		if (dm_request_for_irq_injection(vcpu)) {
-			r = -EINTR;
-			vcpu->run->exit_reason = KVM_EXIT_INTR;
+			r = 0;
+			vcpu->run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
 			++vcpu->stat.request_irq_exits;
 			break;
 		}
-- 
1.8.3.1

[FYI PATCH 06/20] KVM: x86: Split the APIC from the rest of IRQCHIP.

[Paolo Bonzini]

From: Steve Rutherford <srutherford@google.com>

First patch in a series which enables the relocation of the
PIC/IOAPIC to userspace.

Adds capability KVM_CAP_SPLIT_IRQCHIP;

KVM_CAP_SPLIT_IRQCHIP enables the construction of LAPICs without the
rest of the irqchip.

Compile tested for x86.

Signed-off-by: Steve Rutherford <srutherford@google.com>
Suggested-by: Andrew Honig <ahonig@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 Documentation/virtual/kvm/api.txt | 17 +++++++++++++++++
 arch/x86/include/asm/kvm_host.h   |  2 ++
 arch/x86/kvm/i8254.c              |  4 +++-
 arch/x86/kvm/ioapic.h             |  8 ++++++++
 arch/x86/kvm/irq.h                | 11 ++++++++++-
 arch/x86/kvm/irq_comm.c           |  9 ++++++++-
 arch/x86/kvm/lapic.c              |  6 ++++--
 arch/x86/kvm/x86.c                | 23 +++++++++++++++++++++--
 include/linux/kvm_host.h          |  1 +
 include/uapi/linux/kvm.h          |  1 +
 10 files changed, 75 insertions(+), 7 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index d9ecceea5a02..43e0816d0de1 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -3627,6 +3627,23 @@ struct {
 
 KVM handlers should exit to userspace with rc = -EREMOTE.
 
+7.5 KVM_CAP_SPLIT_IRQCHIP
+
+Architectures: x86
+Parameters: None
+Returns: 0 on success, -1 on error
+
+Create a local apic for each processor in the kernel. This can be used
+instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the
+IOAPIC and PIC (and also the PIT, even though this has to be enabled
+separately).
+
+This supersedes KVM_CREATE_IRQCHIP, creating only local APICs, but no in kernel
+IOAPIC or PIC. This also enables in kernel routing of interrupt requests.
+
+Fails if VCPU has already been created, or if the irqchip is already in the
+kernel (i.e. KVM_CREATE_IRQCHIP has already been called).
+
 
 8. Other capabilities.
 ----------------------
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index a0ef289d5a86..befcf555bddc 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -684,6 +684,8 @@ struct kvm_arch {
 	u32 bsp_vcpu_id;
 
 	u64 disabled_quirks;
+
+	bool irqchip_split;
 };
 
 struct kvm_vm_stat {
diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c
index f90952f64e79..08116ff227cc 100644
--- a/arch/x86/kvm/i8254.c
+++ b/arch/x86/kvm/i8254.c
@@ -35,6 +35,7 @@
 #include <linux/kvm_host.h>
 #include <linux/slab.h>
 
+#include "ioapic.h"
 #include "irq.h"
 #include "i8254.h"
 #include "x86.h"
@@ -333,7 +334,8 @@ static void create_pit_timer(struct kvm *kvm, u32 val, int is_period)
 	struct kvm_kpit_state *ps = &kvm->arch.vpit->pit_state;
 	s64 interval;
 
-	if (!irqchip_in_kernel(kvm) || ps->flags & KVM_PIT_FLAGS_HPET_LEGACY)
+	if (!ioapic_in_kernel(kvm) ||
+	    ps->flags & KVM_PIT_FLAGS_HPET_LEGACY)
 		return;
 
 	interval = muldiv64(val, NSEC_PER_SEC, KVM_PIT_FREQ);
diff --git a/arch/x86/kvm/ioapic.h b/arch/x86/kvm/ioapic.h
index bf36d66a1951..a8842c0dee73 100644
--- a/arch/x86/kvm/ioapic.h
+++ b/arch/x86/kvm/ioapic.h
@@ -97,6 +97,14 @@ static inline struct kvm_ioapic *ioapic_irqchip(struct kvm *kvm)
 	return kvm->arch.vioapic;
 }
 
+static inline int ioapic_in_kernel(struct kvm *kvm)
+{
+	int ret;
+
+	ret = (ioapic_irqchip(kvm) != NULL);
+	return ret;
+}
+
 void kvm_rtc_eoi_tracking_restore_one(struct kvm_vcpu *vcpu);
 bool kvm_apic_match_dest(struct kvm_vcpu *vcpu, struct kvm_lapic *source,
 		int short_hand, unsigned int dest, int dest_mode);
diff --git a/arch/x86/kvm/irq.h b/arch/x86/kvm/irq.h
index 9e6e7e04de98..2f9703dcd913 100644
--- a/arch/x86/kvm/irq.h
+++ b/arch/x86/kvm/irq.h
@@ -83,13 +83,22 @@ static inline struct kvm_pic *pic_irqchip(struct kvm *kvm)
 	return kvm->arch.vpic;
 }
 
+static inline int irqchip_split(struct kvm *kvm)
+{
+	return kvm->arch.irqchip_split;
+}
+
 static inline int irqchip_in_kernel(struct kvm *kvm)
 {
 	struct kvm_pic *vpic = pic_irqchip(kvm);
+	bool ret;
+
+	ret = (vpic != NULL);
+	ret |= irqchip_split(kvm);
 
 	/* Read vpic before kvm->irq_routing.  */
 	smp_rmb();
-	return vpic != NULL;
+	return ret;
 }
 
 static inline int lapic_in_kernel(struct kvm_vcpu *vcpu)
diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c
index 9efff9e5b58c..67f6b62a6814 100644
--- a/arch/x86/kvm/irq_comm.c
+++ b/arch/x86/kvm/irq_comm.c
@@ -208,7 +208,7 @@ void kvm_free_irq_source_id(struct kvm *kvm, int irq_source_id)
 		goto unlock;
 	}
 	clear_bit(irq_source_id, &kvm->arch.irq_sources_bitmap);
-	if (!irqchip_in_kernel(kvm))
+	if (!ioapic_in_kernel(kvm))
 		goto unlock;
 
 	kvm_ioapic_clear_all(kvm->arch.vioapic, irq_source_id);
@@ -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 c4bcc86d6dc4..e05946c36b87 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 (ioapic_in_kernel(kvm))
+		kvm_vcpu_request_scan_ioapic(kvm);
 }
 
 static inline void apic_set_spiv(struct kvm_lapic *apic, u32 val)
@@ -1845,7 +1846,8 @@ 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 (ioapic_in_kernel(vcpu->kvm))
+		kvm_rtc_eoi_tracking_restore_one(vcpu);
 }
 
 void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 080aaa7a4d91..b6fbda026708 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2452,6 +2452,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 	case KVM_CAP_ENABLE_CAP_VM:
 	case KVM_CAP_DISABLE_QUIRKS:
 	case KVM_CAP_SET_BOOT_CPU_ID:
+ 	case KVM_CAP_SPLIT_IRQCHIP:
 #ifdef CONFIG_KVM_DEVICE_ASSIGNMENT
 	case KVM_CAP_ASSIGN_DEV_IRQ:
 	case KVM_CAP_PCI_2_3:
@@ -3559,6 +3560,24 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
 		kvm->arch.disabled_quirks = cap->args[0];
 		r = 0;
 		break;
+	case KVM_CAP_SPLIT_IRQCHIP: {
+		mutex_lock(&kvm->lock);
+		r = -EEXIST;
+		if (irqchip_in_kernel(kvm))
+			goto split_irqchip_unlock;
+		if (atomic_read(&kvm->online_vcpus))
+			goto split_irqchip_unlock;
+		r = kvm_setup_empty_irq_routing(kvm);
+		if (r)
+			goto split_irqchip_unlock;
+		/* Pairs with irqchip_in_kernel. */
+		smp_wmb();
+		kvm->arch.irqchip_split = true;
+		r = 0;
+split_irqchip_unlock:
+		mutex_unlock(&kvm->lock);
+		break;
+	}
 	default:
 		r = -EINVAL;
 		break;
@@ -3672,7 +3691,7 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		}
 
 		r = -ENXIO;
-		if (!irqchip_in_kernel(kvm))
+		if (!irqchip_in_kernel(kvm) || irqchip_split(kvm))
 			goto get_irqchip_out;
 		r = kvm_vm_ioctl_get_irqchip(kvm, chip);
 		if (r)
@@ -3696,7 +3715,7 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		}
 
 		r = -ENXIO;
-		if (!irqchip_in_kernel(kvm))
+		if (!irqchip_in_kernel(kvm) || irqchip_split(kvm))
 			goto set_irqchip_out;
 		r = kvm_vm_ioctl_set_irqchip(kvm, chip);
 		if (r)
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 1bef9e21e725..354f147647ab 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -1002,6 +1002,7 @@ 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 a9256f0331ae..ed00f8fc9ea2 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -824,6 +824,7 @@ struct kvm_ppc_smmu_info {
 #define KVM_CAP_MULTI_ADDRESS_SPACE 118
 #define KVM_CAP_GUEST_DEBUG_HW_BPS 119
 #define KVM_CAP_GUEST_DEBUG_HW_WPS 120
+#define KVM_CAP_SPLIT_IRQCHIP 121
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
-- 
1.8.3.1

[FYI PATCH 07/20] KVM: x86: Add KVM exit for IOAPIC EOIs

[Paolo Bonzini]

From: Steve Rutherford <srutherford@google.com>

Adds KVM_EXIT_IOAPIC_EOI which allows the kernel to EOI
level-triggered IOAPIC interrupts.

Uses a per VCPU exit bitmap to decide whether or not the IOAPIC needs
to be informed (which is identical to the EOI_EXIT_BITMAP field used
by modern x86 processors, but can also be used to elide kvm IOAPIC EOI
exits on older processors).

[Note: A prototype using ResampleFDs found that decoupling the EOI
from the VCPU's thread made it possible for the VCPU to not see a
recent EOI after reentering the guest. This does not match real
hardware.]

Compile tested for Intel x86.

Signed-off-by: Steve Rutherford <srutherford@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 Documentation/virtual/kvm/api.txt | 12 ++++++++++++
 arch/x86/include/asm/kvm_host.h   |  2 ++
 arch/x86/kvm/lapic.c              | 24 +++++++++++++++++-------
 arch/x86/kvm/x86.c                | 11 +++++++++++
 include/linux/kvm_host.h          |  2 +-
 include/uapi/linux/kvm.h          |  5 +++++
 6 files changed, 48 insertions(+), 8 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 43e0816d0de1..0d14bf5db534 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -3309,6 +3309,18 @@ Valid values for 'type' are:
    to ignore the request, or to gather VM memory core dump and/or
    reset/shutdown of the VM.
 
+		/* KVM_EXIT_IOAPIC_EOI */
+		struct {
+			__u8 vector;
+		} eoi;
+
+Indicates that the VCPU's in-kernel local APIC received an EOI for a
+level-triggered IOAPIC interrupt.  This exit only triggers when the
+IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled);
+the userspace IOAPIC should process the EOI and retrigger the interrupt if
+it is still asserted.  Vector is the LAPIC interrupt vector for which the
+EOI was received.
+
 		/* Fix the size of the union. */
 		char padding[256];
 	};
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index befcf555bddc..af09fa1d1be7 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -574,6 +574,8 @@ struct kvm_vcpu_arch {
 	struct {
 		bool pv_unhalted;
 	} pv;
+
+	int pending_ioapic_eoi;
 };
 
 struct kvm_lpage_info {
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index e05946c36b87..ef70f6f3a37a 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -877,15 +877,25 @@ static bool kvm_ioapic_handles_vector(struct kvm_lapic *apic, int vector)
 
 static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector)
 {
-	if (kvm_ioapic_handles_vector(apic, vector)) {
-		int trigger_mode;
-		if (apic_test_vector(vector, apic->regs + APIC_TMR))
-			trigger_mode = IOAPIC_LEVEL_TRIG;
-		else
-			trigger_mode = IOAPIC_EDGE_TRIG;
+	int trigger_mode;
+
+	/* Eoi the ioapic only if the ioapic doesn't own the vector. */
+	if (!kvm_ioapic_handles_vector(apic, vector))
+		return;
 
-		kvm_ioapic_update_eoi(apic->vcpu, vector, trigger_mode);
+	/* Request a KVM exit to inform the userspace IOAPIC. */
+	if (irqchip_split(apic->vcpu->kvm)) {
+		apic->vcpu->arch.pending_ioapic_eoi = vector;
+		kvm_make_request(KVM_REQ_IOAPIC_EOI_EXIT, apic->vcpu);
+		return;
 	}
+
+	if (apic_test_vector(vector, apic->regs + APIC_TMR))
+		trigger_mode = IOAPIC_LEVEL_TRIG;
+	else
+		trigger_mode = IOAPIC_EDGE_TRIG;
+
+	kvm_ioapic_update_eoi(apic->vcpu, vector, trigger_mode);
 }
 
 static int apic_set_eoi(struct kvm_lapic *apic)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index b6fbda026708..07604e205c99 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6275,6 +6275,17 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 			kvm_pmu_handle_event(vcpu);
 		if (kvm_check_request(KVM_REQ_PMI, vcpu))
 			kvm_pmu_deliver_pmi(vcpu);
+		if (kvm_check_request(KVM_REQ_IOAPIC_EOI_EXIT, vcpu)) {
+			BUG_ON(vcpu->arch.pending_ioapic_eoi > 255);
+			if (test_bit(vcpu->arch.pending_ioapic_eoi,
+				     (void *) vcpu->arch.eoi_exit_bitmap)) {
+				vcpu->run->exit_reason = KVM_EXIT_IOAPIC_EOI;
+				vcpu->run->eoi.vector =
+						vcpu->arch.pending_ioapic_eoi;
+				r = 0;
+				goto out;
+			}
+		}
 		if (kvm_check_request(KVM_REQ_SCAN_IOAPIC, vcpu))
 			vcpu_scan_ioapic(vcpu);
 		if (kvm_check_request(KVM_REQ_APIC_PAGE_RELOAD, vcpu))
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 354f147647ab..3b33215ed447 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -140,6 +140,7 @@ static inline bool is_error_page(struct page *page)
 #define KVM_REQ_APIC_PAGE_RELOAD  25
 #define KVM_REQ_SMI               26
 #define KVM_REQ_HV_CRASH          27
+#define KVM_REQ_IOAPIC_EOI_EXIT   28
 
 #define KVM_USERSPACE_IRQ_SOURCE_ID		0
 #define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
@@ -1146,4 +1147,3 @@ static inline void kvm_vcpu_set_dy_eligible(struct kvm_vcpu *vcpu, bool val)
 }
 #endif /* CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT */
 #endif
-
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index ed00f8fc9ea2..12e3afbf0f47 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -183,6 +183,7 @@ struct kvm_s390_skeys {
 #define KVM_EXIT_EPR              23
 #define KVM_EXIT_SYSTEM_EVENT     24
 #define KVM_EXIT_S390_STSI        25
+#define KVM_EXIT_IOAPIC_EOI       26
 
 /* For KVM_EXIT_INTERNAL_ERROR */
 /* Emulate instruction failed. */
@@ -333,6 +334,10 @@ struct kvm_run {
 			__u8 sel1;
 			__u16 sel2;
 		} s390_stsi;
+		/* KVM_EXIT_IOAPIC_EOI */
+		struct {
+			__u8 vector;
+		} eoi;
 		/* Fix the size of the union. */
 		char padding[256];
 	};
-- 
1.8.3.1

[FYI PATCH 08/20] KVM: x86: Add EOI exit bitmap inference

[Paolo Bonzini]

From: Steve Rutherford <srutherford@google.com>

In order to support a userspace IOAPIC interacting with an in kernel
APIC, the EOI exit bitmaps need to be configurable.

If the IOAPIC is in userspace (i.e. the irqchip has been split), the
EOI exit bitmaps will be set whenever the GSI Routes are configured.
In particular, for the low MSI routes are reservable for userspace
IOAPICs. For these MSI routes, the EOI Exit bit corresponding to the
destination vector of the route will be set for the destination VCPU.

The intention is for the userspace IOAPICs to use the reservable MSI
routes to inject interrupts into the guest.

This is a slight abuse of the notion of an MSI Route, given that MSIs
classically bypass the IOAPIC. It might be worthwhile to add an
additional route type to improve clarity.

Compile tested for Intel x86.

Signed-off-by: Steve Rutherford <srutherford@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 Documentation/virtual/kvm/api.txt |  9 ++++++---
 arch/x86/include/asm/kvm_host.h   |  1 +
 arch/x86/kvm/ioapic.h             |  2 ++
 arch/x86/kvm/irq_comm.c           | 42 +++++++++++++++++++++++++++++++++++++++
 arch/x86/kvm/lapic.c              |  3 +--
 arch/x86/kvm/x86.c                |  9 ++++++++-
 include/linux/kvm_host.h          | 17 ++++++++++++++++
 virt/kvm/irqchip.c                | 12 ++---------
 8 files changed, 79 insertions(+), 16 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 0d14bf5db534..89e71648d748 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -3642,7 +3642,7 @@ KVM handlers should exit to userspace with rc = -EREMOTE.
 7.5 KVM_CAP_SPLIT_IRQCHIP
 
 Architectures: x86
-Parameters: None
+Parameters: args[0] - number of routes reserved for userspace IOAPICs
 Returns: 0 on success, -1 on error
 
 Create a local apic for each processor in the kernel. This can be used
@@ -3650,8 +3650,11 @@ instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the
 IOAPIC and PIC (and also the PIT, even though this has to be enabled
 separately).
 
-This supersedes KVM_CREATE_IRQCHIP, creating only local APICs, but no in kernel
-IOAPIC or PIC. This also enables in kernel routing of interrupt requests.
+This capability also enables in kernel routing of interrupt requests;
+when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are
+used in the IRQ routing table.  The first args[0] MSI routes are reserved
+for the IOAPIC pins.  Whenever the LAPIC receives an EOI for these routes,
+a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace.
 
 Fails if VCPU has already been created, or if the irqchip is already in the
 kernel (i.e. KVM_CREATE_IRQCHIP has already been called).
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index af09fa1d1be7..7a5f9debbcd8 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -688,6 +688,7 @@ struct kvm_arch {
 	u64 disabled_quirks;
 
 	bool irqchip_split;
+	u8 nr_reserved_ioapic_pins;
 };
 
 struct kvm_vm_stat {
diff --git a/arch/x86/kvm/ioapic.h b/arch/x86/kvm/ioapic.h
index a8842c0dee73..084617d37c74 100644
--- a/arch/x86/kvm/ioapic.h
+++ b/arch/x86/kvm/ioapic.h
@@ -9,6 +9,7 @@ struct kvm;
 struct kvm_vcpu;
 
 #define IOAPIC_NUM_PINS  KVM_IOAPIC_NUM_PINS
+#define MAX_NR_RESERVED_IOAPIC_PINS KVM_MAX_IRQ_ROUTES
 #define IOAPIC_VERSION_ID 0x11	/* IOAPIC version */
 #define IOAPIC_EDGE_TRIG  0
 #define IOAPIC_LEVEL_TRIG 1
@@ -121,5 +122,6 @@ int kvm_irq_delivery_to_apic(struct kvm *kvm, struct kvm_lapic *src,
 int kvm_get_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
 int kvm_set_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
 void kvm_ioapic_scan_entry(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap);
+void kvm_scan_ioapic_routes(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap);
 
 #endif
diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c
index 67f6b62a6814..177460998bb0 100644
--- a/arch/x86/kvm/irq_comm.c
+++ b/arch/x86/kvm/irq_comm.c
@@ -335,3 +335,45 @@ int kvm_setup_empty_irq_routing(struct kvm *kvm)
 {
 	return kvm_set_irq_routing(kvm, empty_routing, 0, 0);
 }
+
+void kvm_arch_irq_routing_update(struct kvm *kvm)
+{
+	if (ioapic_in_kernel(kvm) || !irqchip_in_kernel(kvm))
+		return;
+	kvm_make_scan_ioapic_request(kvm);
+}
+
+void kvm_scan_ioapic_routes(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
+{
+	struct kvm *kvm = vcpu->kvm;
+	struct kvm_kernel_irq_routing_entry *entry;
+	struct kvm_irq_routing_table *table;
+	u32 i, nr_ioapic_pins;
+	int idx;
+
+	/* kvm->irq_routing must be read after clearing
+	 * KVM_SCAN_IOAPIC. */
+	smp_mb();
+	idx = srcu_read_lock(&kvm->irq_srcu);
+	table = srcu_dereference(kvm->irq_routing, &kvm->irq_srcu);
+	nr_ioapic_pins = min_t(u32, table->nr_rt_entries,
+			       kvm->arch.nr_reserved_ioapic_pins);
+	for (i = 0; i < nr_ioapic_pins; ++i) {
+		hlist_for_each_entry(entry, &table->map[i], link) {
+			u32 dest_id, dest_mode;
+
+			if (entry->type != KVM_IRQ_ROUTING_MSI)
+				continue;
+			dest_id = (entry->msi.address_lo >> 12) & 0xff;
+			dest_mode = (entry->msi.address_lo >> 2) & 0x1;
+			if (kvm_apic_match_dest(vcpu, NULL, 0, dest_id,
+						dest_mode)) {
+				u32 vector = entry->msi.data & 0xff;
+
+				__set_bit(vector,
+					  (unsigned long *) eoi_exit_bitmap);
+			}
+		}
+	}
+	srcu_read_unlock(&kvm->irq_srcu, idx);
+}
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index ef70f6f3a37a..2f4c0d0cbe0a 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -209,8 +209,7 @@ out:
 	if (old)
 		kfree_rcu(old, rcu);
 
-	if (ioapic_in_kernel(kvm))
-		kvm_vcpu_request_scan_ioapic(kvm);
+	kvm_make_scan_ioapic_request(kvm);
 }
 
 static inline void apic_set_spiv(struct kvm_lapic *apic, u32 val)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 07604e205c99..99ebc8e02097 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3562,6 +3562,9 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
 		break;
 	case KVM_CAP_SPLIT_IRQCHIP: {
 		mutex_lock(&kvm->lock);
+		r = -EINVAL;
+		if (cap->args[0] > MAX_NR_RESERVED_IOAPIC_PINS)
+			goto split_irqchip_unlock;
 		r = -EEXIST;
 		if (irqchip_in_kernel(kvm))
 			goto split_irqchip_unlock;
@@ -3573,6 +3576,7 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
 		/* Pairs with irqchip_in_kernel. */
 		smp_wmb();
 		kvm->arch.irqchip_split = true;
+		kvm->arch.nr_reserved_ioapic_pins = cap->args[0];
 		r = 0;
 split_irqchip_unlock:
 		mutex_unlock(&kvm->lock);
@@ -6171,7 +6175,10 @@ static void vcpu_scan_ioapic(struct kvm_vcpu *vcpu)
 
 	memset(vcpu->arch.eoi_exit_bitmap, 0, 256 / 8);
 
-	kvm_ioapic_scan_entry(vcpu, vcpu->arch.eoi_exit_bitmap);
+	if (irqchip_split(vcpu->kvm))
+		kvm_scan_ioapic_routes(vcpu, vcpu->arch.eoi_exit_bitmap);
+	else
+		kvm_ioapic_scan_entry(vcpu, vcpu->arch.eoi_exit_bitmap);
 	kvm_x86_ops->load_eoi_exitmap(vcpu);
 }
 
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 3b33215ed447..a73d60d474dd 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -330,6 +330,19 @@ struct kvm_kernel_irq_routing_entry {
 	struct hlist_node link;
 };
 
+#ifdef CONFIG_HAVE_KVM_IRQCHIP
+struct kvm_irq_routing_table {
+	int chip[KVM_NR_IRQCHIPS][KVM_IRQCHIP_NUM_PINS];
+	struct kvm_kernel_irq_routing_entry *rt_entries;
+	u32 nr_rt_entries;
+	/*
+	 * Array indexed by gsi. Each entry contains list of irq chips
+	 * the gsi is connected to.
+	 */
+	struct hlist_head map[0];
+};
+#endif
+
 #ifndef KVM_PRIVATE_MEM_SLOTS
 #define KVM_PRIVATE_MEM_SLOTS 0
 #endif
@@ -456,10 +469,14 @@ void vcpu_put(struct kvm_vcpu *vcpu);
 
 #ifdef __KVM_HAVE_IOAPIC
 void kvm_vcpu_request_scan_ioapic(struct kvm *kvm);
+void kvm_arch_irq_routing_update(struct kvm *kvm);
 #else
 static inline void kvm_vcpu_request_scan_ioapic(struct kvm *kvm)
 {
 }
+static inline void kvm_arch_irq_routing_update(struct kvm *kvm)
+{
+}
 #endif
 
 #ifdef CONFIG_HAVE_KVM_IRQFD
diff --git a/virt/kvm/irqchip.c b/virt/kvm/irqchip.c
index d7ea8e20dae4..716a1c4db528 100644
--- a/virt/kvm/irqchip.c
+++ b/virt/kvm/irqchip.c
@@ -31,16 +31,6 @@
 #include <trace/events/kvm.h>
 #include "irq.h"
 
-struct kvm_irq_routing_table {
-	int chip[KVM_NR_IRQCHIPS][KVM_IRQCHIP_NUM_PINS];
-	u32 nr_rt_entries;
-	/*
-	 * Array indexed by gsi. Each entry contains list of irq chips
-	 * the gsi is connected to.
-	 */
-	struct hlist_head map[0];
-};
-
 int kvm_irq_map_gsi(struct kvm *kvm,
 		    struct kvm_kernel_irq_routing_entry *entries, int gsi)
 {
@@ -231,6 +221,8 @@ int kvm_set_irq_routing(struct kvm *kvm,
 	kvm_irq_routing_update(kvm);
 	mutex_unlock(&kvm->irq_lock);
 
+	kvm_arch_irq_routing_update(kvm);
+
 	synchronize_srcu_expedited(&kvm->irq_srcu);
 
 	new = old;
-- 
1.8.3.1

[FYI PATCH 09/20] KVM: x86: Add support for local interrupt requests from userspace

[Paolo Bonzini]

From: Steve Rutherford <srutherford@google.com>

In order to enable userspace PIC support, the userspace PIC needs to
be able to inject local interrupts even when the APICs are in the
kernel.

KVM_INTERRUPT now supports sending local interrupts to an APIC when
APICs are in the kernel.

The ready_for_interrupt_request flag is now only set when the CPU/APIC
will immediately accept and inject an interrupt (i.e. APIC has not
masked the PIC).

When the PIC wishes to initiate an INTA cycle with, say, CPU0, it
kicks CPU0 out of the guest, and renedezvous with CPU0 once it arrives
in userspace.

When the CPU/APIC unmasks the PIC, a KVM_EXIT_IRQ_WINDOW_OPEN is
triggered, so that userspace has a chance to inject a PIC interrupt
if it had been pending.

Overall, this design can lead to a small number of spurious userspace
renedezvous. In particular, whenever the PIC transistions from low to
high while it is masked and whenever the PIC becomes unmasked while
it is low.

Note: this does not buffer more than one local interrupt in the
kernel, so the VMM needs to enter the guest in order to complete
interrupt injection before injecting an additional interrupt.

Compiles for x86.

Can pass the KVM Unit Tests.

Signed-off-by: Steve Rutherford <srutherford@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 Documentation/virtual/kvm/api.txt | 14 +++++++++----
 arch/x86/include/asm/kvm_host.h   |  1 +
 arch/x86/kvm/irq.c                | 32 +++++++++++++++++++++++------
 arch/x86/kvm/irq.h                |  8 ++++++++
 arch/x86/kvm/x86.c                | 42 ++++++++++++++++++++++++++++++---------
 5 files changed, 78 insertions(+), 19 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 89e71648d748..e3e9c41721a2 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -401,10 +401,9 @@ Capability: basic
 Architectures: x86, ppc, mips
 Type: vcpu ioctl
 Parameters: struct kvm_interrupt (in)
-Returns: 0 on success, -1 on error
+Returns: 0 on success, negative on failure.
 
-Queues a hardware interrupt vector to be injected.  This is only
-useful if in-kernel local APIC or equivalent is not used.
+Queues a hardware interrupt vector to be injected.
 
 /* for KVM_INTERRUPT */
 struct kvm_interrupt {
@@ -414,7 +413,14 @@ struct kvm_interrupt {
 
 X86:
 
-Note 'irq' is an interrupt vector, not an interrupt pin or line.
+Returns: 0 on success,
+	 -EEXIST if an interrupt is already enqueued
+	 -EINVAL the the irq number is invalid
+	 -ENXIO if the PIC is in the kernel
+	 -EFAULT if the pointer is invalid
+
+Note 'irq' is an interrupt vector, not an interrupt pin or line. This
+ioctl is useful if the in-kernel PIC is not used.
 
 PPC:
 
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 7a5f9debbcd8..76a5b30979b3 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -576,6 +576,7 @@ struct kvm_vcpu_arch {
 	} pv;
 
 	int pending_ioapic_eoi;
+	int pending_external_vector;
 };
 
 struct kvm_lpage_info {
diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c
index b653ae202c8e..097060e33bd6 100644
--- a/arch/x86/kvm/irq.c
+++ b/arch/x86/kvm/irq.c
@@ -38,14 +38,27 @@ int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
 EXPORT_SYMBOL(kvm_cpu_has_pending_timer);
 
 /*
+ * check if there is a pending userspace external interrupt
+ */
+static int pending_userspace_extint(struct kvm_vcpu *v)
+{
+	return v->arch.pending_external_vector != -1;
+}
+
+/*
  * check if there is pending interrupt from
  * non-APIC source without intack.
  */
 static int kvm_cpu_has_extint(struct kvm_vcpu *v)
 {
-	if (kvm_apic_accept_pic_intr(v))
-		return pic_irqchip(v->kvm)->output;	/* PIC */
-	else
+	u8 accept = kvm_apic_accept_pic_intr(v);
+
+	if (accept) {
+		if (irqchip_split(v->kvm))
+			return pending_userspace_extint(v);
+		else
+			return pic_irqchip(v->kvm)->output;
+	} else
 		return 0;
 }
 
@@ -91,9 +104,16 @@ EXPORT_SYMBOL_GPL(kvm_cpu_has_interrupt);
  */
 static int kvm_cpu_get_extint(struct kvm_vcpu *v)
 {
-	if (kvm_cpu_has_extint(v))
-		return kvm_pic_read_irq(v->kvm); /* PIC */
-	return -1;
+	if (kvm_cpu_has_extint(v)) {
+		if (irqchip_split(v->kvm)) {
+			int vector = v->arch.pending_external_vector;
+
+			v->arch.pending_external_vector = -1;
+			return vector;
+		} else
+			return kvm_pic_read_irq(v->kvm); /* PIC */
+	} else
+		return -1;
 }
 
 /*
diff --git a/arch/x86/kvm/irq.h b/arch/x86/kvm/irq.h
index 2f9703dcd913..ae5c78f2337d 100644
--- a/arch/x86/kvm/irq.h
+++ b/arch/x86/kvm/irq.h
@@ -83,6 +83,14 @@ static inline struct kvm_pic *pic_irqchip(struct kvm *kvm)
 	return kvm->arch.vpic;
 }
 
+static inline int pic_in_kernel(struct kvm *kvm)
+{
+	int ret;
+
+	ret = (pic_irqchip(kvm) != NULL);
+	return ret;
+}
+
 static inline int irqchip_split(struct kvm *kvm)
 {
 	return kvm->arch.irqchip_split;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 99ebc8e02097..1264748e1cc4 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2666,12 +2666,24 @@ 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 (!irqchip_in_kernel(vcpu->kvm)) {
+		kvm_queue_interrupt(vcpu, irq->irq, false);
+		kvm_make_request(KVM_REQ_EVENT, vcpu);
+		return 0;
+	}
+
+	/*
+	 * With in-kernel LAPIC, we only use this to inject EXTINT, so
+	 * fail for in-kernel 8259.
+	 */
+	if (pic_in_kernel(vcpu->kvm))
 		return -ENXIO;
 
-	kvm_queue_interrupt(vcpu, irq->irq, false);
-	kvm_make_request(KVM_REQ_EVENT, vcpu);
+	if (vcpu->arch.pending_external_vector != -1)
+		return -EEXIST;
 
+	vcpu->arch.pending_external_vector = irq->irq;
 	return 0;
 }
 
@@ -5800,9 +5812,15 @@ 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) &&
-		vcpu->run->request_interrupt_window &&
-		kvm_arch_interrupt_allowed(vcpu));
+	if (!vcpu->run->request_interrupt_window || pic_in_kernel(vcpu->kvm))
+		return false;
+
+	if (kvm_cpu_has_interrupt(vcpu))
+		return false;
+
+	return (irqchip_split(vcpu->kvm)
+		? kvm_apic_accept_pic_intr(vcpu)
+		: kvm_arch_interrupt_allowed(vcpu));
 }
 
 static void post_kvm_run_save(struct kvm_vcpu *vcpu)
@@ -5813,13 +5831,17 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu)
 	kvm_run->flags = is_smm(vcpu) ? KVM_RUN_X86_SMM : 0;
 	kvm_run->cr8 = kvm_get_cr8(vcpu);
 	kvm_run->apic_base = kvm_get_apic_base(vcpu);
-	if (irqchip_in_kernel(vcpu->kvm))
-		kvm_run->ready_for_interrupt_injection = 1;
-	else
+	if (!irqchip_in_kernel(vcpu->kvm))
 		kvm_run->ready_for_interrupt_injection =
 			kvm_arch_interrupt_allowed(vcpu) &&
 			!kvm_cpu_has_interrupt(vcpu) &&
 			!kvm_event_needs_reinjection(vcpu);
+	else if (!pic_in_kernel(vcpu->kvm))
+		kvm_run->ready_for_interrupt_injection =
+			kvm_apic_accept_pic_intr(vcpu) &&
+			!kvm_cpu_has_interrupt(vcpu);
+	else
+		kvm_run->ready_for_interrupt_injection = 1;
 }
 
 static void update_cr8_intercept(struct kvm_vcpu *vcpu)
@@ -7407,6 +7429,8 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 	kvm_async_pf_hash_reset(vcpu);
 	kvm_pmu_init(vcpu);
 
+	vcpu->arch.pending_external_vector = -1;
+
 	return 0;
 
 fail_free_mce_banks:
-- 
1.8.3.1

[FYI PATCH 10/20] kvm: use kmalloc() instead of kzalloc() during iodev register/unregister

[Paolo Bonzini]

From: Jason Wang <jasowang@redhat.com>

All fields of kvm_io_range were initialized or copied explicitly
afterwards. So switch to use kmalloc().

Cc: Gleb Natapov <gleb@kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 virt/kvm/kvm_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 8db1d9361993..23116dcb2129 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -3341,7 +3341,7 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
 	if (bus->dev_count - bus->ioeventfd_count > NR_IOBUS_DEVS - 1)
 		return -ENOSPC;
 
-	new_bus = kzalloc(sizeof(*bus) + ((bus->dev_count + 1) *
+	new_bus = kmalloc(sizeof(*bus) + ((bus->dev_count + 1) *
 			  sizeof(struct kvm_io_range)), GFP_KERNEL);
 	if (!new_bus)
 		return -ENOMEM;
@@ -3373,7 +3373,7 @@ int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
 	if (r)
 		return r;
 
-	new_bus = kzalloc(sizeof(*bus) + ((bus->dev_count - 1) *
+	new_bus = kmalloc(sizeof(*bus) + ((bus->dev_count - 1) *
 			  sizeof(struct kvm_io_range)), GFP_KERNEL);
 	if (!new_bus)
 		return -ENOMEM;
-- 
1.8.3.1

[FYI PATCH 11/20] kvm: add tracepoint for fast mmio

[Paolo Bonzini]

From: Jason Wang <jasowang@redhat.com>

Cc: Gleb Natapov <gleb@kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/trace.h | 18 ++++++++++++++++++
 arch/x86/kvm/vmx.c   |  1 +
 arch/x86/kvm/x86.c   |  1 +
 3 files changed, 20 insertions(+)

diff --git a/arch/x86/kvm/trace.h b/arch/x86/kvm/trace.h
index 4eae7c35ddf5..ce4abe333c39 100644
--- a/arch/x86/kvm/trace.h
+++ b/arch/x86/kvm/trace.h
@@ -129,6 +129,24 @@ TRACE_EVENT(kvm_pio,
 );
 
 /*
+ * Tracepoint for fast mmio.
+ */
+TRACE_EVENT(kvm_fast_mmio,
+	TP_PROTO(u64 gpa),
+	TP_ARGS(gpa),
+
+	TP_STRUCT__entry(
+		__field(u64,	gpa)
+	),
+
+	TP_fast_assign(
+		__entry->gpa		= gpa;
+	),
+
+	TP_printk("fast mmio at gpa 0x%llx", __entry->gpa)
+);
+
+/*
  * Tracepoint for cpuid.
  */
 TRACE_EVENT(kvm_cpuid,
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 90605f70a7d0..b2d619ab3d07 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -5756,6 +5756,7 @@ static int handle_ept_misconfig(struct kvm_vcpu *vcpu)
 	gpa = vmcs_read64(GUEST_PHYSICAL_ADDRESS);
 	if (!kvm_io_bus_write(vcpu, KVM_FAST_MMIO_BUS, gpa, 0, NULL)) {
 		skip_emulated_instruction(vcpu);
+		trace_kvm_fast_mmio(gpa);
 		return 1;
 	}
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 1264748e1cc4..b015ab357980 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -8074,6 +8074,7 @@ bool kvm_arch_has_noncoherent_dma(struct kvm *kvm)
 EXPORT_SYMBOL_GPL(kvm_arch_has_noncoherent_dma);
 
 EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_exit);
+EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_fast_mmio);
 EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_inj_virq);
 EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_page_fault);
 EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_msr);
-- 
1.8.3.1

[FYI PATCH 12/20] kvm: add capability for any-length ioeventfds

[Paolo Bonzini]

From: Jason Wang <jasowang@redhat.com>

Cc: Gleb Natapov <gleb@kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 Documentation/virtual/kvm/api.txt | 6 +++++-
 include/uapi/linux/kvm.h          | 1 +
 virt/kvm/eventfd.c                | 4 +---
 virt/kvm/kvm_main.c               | 1 +
 4 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index e3e9c41721a2..34cc068e81ea 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -1604,7 +1604,7 @@ provided event instead of triggering an exit.
 struct kvm_ioeventfd {
 	__u64 datamatch;
 	__u64 addr;        /* legal pio/mmio address */
-	__u32 len;         /* 1, 2, 4, or 8 bytes    */
+	__u32 len;         /* 0, 1, 2, 4, or 8 bytes    */
 	__s32 fd;
 	__u32 flags;
 	__u8  pad[36];
@@ -1627,6 +1627,10 @@ to the registered address is equal to datamatch in struct kvm_ioeventfd.
 For virtio-ccw devices, addr contains the subchannel id and datamatch the
 virtqueue index.
 
+With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and
+the kernel will ignore the length of guest write and may get a faster vmexit.
+The speedup may only apply to specific architectures, but the ioeventfd will
+work anyway.
 
 4.60 KVM_DIRTY_TLB
 
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index 12e3afbf0f47..03f3618612aa 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -830,6 +830,7 @@ struct kvm_ppc_smmu_info {
 #define KVM_CAP_GUEST_DEBUG_HW_BPS 119
 #define KVM_CAP_GUEST_DEBUG_HW_WPS 120
 #define KVM_CAP_SPLIT_IRQCHIP 121
+#define KVM_CAP_IOEVENTFD_ANY_LENGTH 122
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c
index 79db45336e3a..ac89299b8699 100644
--- a/virt/kvm/eventfd.c
+++ b/virt/kvm/eventfd.c
@@ -914,9 +914,7 @@ kvm_assign_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
 		return -EINVAL;
 
 	/* ioeventfd with no length can't be combined with DATAMATCH */
-	if (!args->len &&
-	    args->flags & (KVM_IOEVENTFD_FLAG_PIO |
-			   KVM_IOEVENTFD_FLAG_DATAMATCH))
+	if (!args->len && (args->flags & KVM_IOEVENTFD_FLAG_DATAMATCH))
 		return -EINVAL;
 
 	ret = kvm_assign_ioeventfd_idx(kvm, bus_idx, args);
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 23116dcb2129..afd7ae6aec65 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2718,6 +2718,7 @@ static long kvm_vm_ioctl_check_extension_generic(struct kvm *kvm, long arg)
 	case KVM_CAP_IRQFD:
 	case KVM_CAP_IRQFD_RESAMPLE:
 #endif
+	case KVM_CAP_IOEVENTFD_ANY_LENGTH:
 	case KVM_CAP_CHECK_EXTENSION_VM:
 		return 1;
 #ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
-- 
1.8.3.1

[FYI PATCH 13/20] kvm/x86: Hyper-V HV_X64_MSR_RESET msr

[Paolo Bonzini]

From: Andrey Smetanin <asmetanin@virtuozzo.com>

HV_X64_MSR_RESET msr is used by Hyper-V based Windows guest
to reset guest VM by hypervisor.

Necessary to support loading of winhv.sys in guest, which in turn is
required to support Windows VMBus.

Signed-off-by: Andrey Smetanin <asmetanin@virtuozzo.com>
Reviewed-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: Gleb Natapov <gleb@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/uapi/asm/hyperv.h |  3 +++
 arch/x86/kvm/hyperv.c              | 10 ++++++++++
 arch/x86/kvm/x86.c                 |  7 +++++++
 include/linux/kvm_host.h           |  1 +
 4 files changed, 21 insertions(+)

diff --git a/arch/x86/include/uapi/asm/hyperv.h b/arch/x86/include/uapi/asm/hyperv.h
index f0412c50c47b..dab584bf7ddf 100644
--- a/arch/x86/include/uapi/asm/hyperv.h
+++ b/arch/x86/include/uapi/asm/hyperv.h
@@ -153,6 +153,9 @@
 /* MSR used to provide vcpu index */
 #define HV_X64_MSR_VP_INDEX			0x40000002
 
+/* MSR used to reset the guest OS. */
+#define HV_X64_MSR_RESET			0x40000003
+
 /* MSR used to read the per-partition time reference counter */
 #define HV_X64_MSR_TIME_REF_COUNT		0x40000020
 
diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
index a8160d2ae362..0ad11a232474 100644
--- a/arch/x86/kvm/hyperv.c
+++ b/arch/x86/kvm/hyperv.c
@@ -41,6 +41,7 @@ static bool kvm_hv_msr_partition_wide(u32 msr)
 	case HV_X64_MSR_TIME_REF_COUNT:
 	case HV_X64_MSR_CRASH_CTL:
 	case HV_X64_MSR_CRASH_P0 ... HV_X64_MSR_CRASH_P4:
+	case HV_X64_MSR_RESET:
 		r = true;
 		break;
 	}
@@ -163,6 +164,12 @@ static int kvm_hv_set_msr_pw(struct kvm_vcpu *vcpu, u32 msr, u64 data,
 						 data);
 	case HV_X64_MSR_CRASH_CTL:
 		return kvm_hv_msr_set_crash_ctl(vcpu, data, host);
+	case HV_X64_MSR_RESET:
+		if (data == 1) {
+			vcpu_debug(vcpu, "hyper-v reset requested\n");
+			kvm_make_request(KVM_REQ_HV_RESET, vcpu);
+		}
+		break;
 	default:
 		vcpu_unimpl(vcpu, "Hyper-V uhandled wrmsr: 0x%x data 0x%llx\n",
 			    msr, data);
@@ -241,6 +248,9 @@ static int kvm_hv_get_msr_pw(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 						 pdata);
 	case HV_X64_MSR_CRASH_CTL:
 		return kvm_hv_msr_get_crash_ctl(vcpu, pdata);
+	case HV_X64_MSR_RESET:
+		data = 0;
+		break;
 	default:
 		vcpu_unimpl(vcpu, "Hyper-V unhandled rdmsr: 0x%x\n", msr);
 		return 1;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index b015ab357980..7e207de0a13f 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -952,6 +952,7 @@ static u32 emulated_msrs[] = {
 	HV_X64_MSR_TIME_REF_COUNT, HV_X64_MSR_REFERENCE_TSC,
 	HV_X64_MSR_CRASH_P0, HV_X64_MSR_CRASH_P1, HV_X64_MSR_CRASH_P2,
 	HV_X64_MSR_CRASH_P3, HV_X64_MSR_CRASH_P4, HV_X64_MSR_CRASH_CTL,
+	HV_X64_MSR_RESET,
 	HV_X64_MSR_APIC_ASSIST_PAGE, MSR_KVM_ASYNC_PF_EN, MSR_KVM_STEAL_TIME,
 	MSR_KVM_PV_EOI_EN,
 
@@ -6325,6 +6326,12 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 			r = 0;
 			goto out;
 		}
+		if (kvm_check_request(KVM_REQ_HV_RESET, vcpu)) {
+			vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
+			vcpu->run->system_event.type = KVM_SYSTEM_EVENT_RESET;
+			r = 0;
+			goto out;
+		}
 	}
 
 	if (kvm_check_request(KVM_REQ_EVENT, vcpu) || req_int_win) {
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index a73d60d474dd..f3bd0fd0795e 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -141,6 +141,7 @@ static inline bool is_error_page(struct page *page)
 #define KVM_REQ_SMI               26
 #define KVM_REQ_HV_CRASH          27
 #define KVM_REQ_IOAPIC_EOI_EXIT   28
+#define KVM_REQ_HV_RESET          29
 
 #define KVM_USERSPACE_IRQ_SOURCE_ID		0
 #define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
-- 
1.8.3.1

[FYI PATCH 14/20] kvm/x86: Hyper-V HV_X64_MSR_VP_INDEX export for QEMU.

[Paolo Bonzini]

From: Andrey Smetanin <asmetanin@virtuozzo.com>

Insert Hyper-V HV_X64_MSR_VP_INDEX into msr's emulated list,
so QEMU can set Hyper-V features cpuid HV_X64_MSR_VP_INDEX_AVAILABLE
bit correctly. KVM emulation part is in place already.

Necessary to support loading of winhv.sys in guest, which in turn is
required to support Windows VMBus.

Signed-off-by: Andrey Smetanin <asmetanin@virtuozzo.com>
Reviewed-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: Gleb Natapov <gleb@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/x86.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 7e207de0a13f..5bc598de28a7 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -953,6 +953,7 @@ static u32 emulated_msrs[] = {
 	HV_X64_MSR_CRASH_P0, HV_X64_MSR_CRASH_P1, HV_X64_MSR_CRASH_P2,
 	HV_X64_MSR_CRASH_P3, HV_X64_MSR_CRASH_P4, HV_X64_MSR_CRASH_CTL,
 	HV_X64_MSR_RESET,
+	HV_X64_MSR_VP_INDEX,
 	HV_X64_MSR_APIC_ASSIST_PAGE, MSR_KVM_ASYNC_PF_EN, MSR_KVM_STEAL_TIME,
 	MSR_KVM_PV_EOI_EN,
 
-- 
1.8.3.1

[FYI PATCH 15/20] kvm/x86: Hyper-V HV_X64_MSR_VP_RUNTIME support

[Paolo Bonzini]

From: Andrey Smetanin <asmetanin@virtuozzo.com>

HV_X64_MSR_VP_RUNTIME msr used by guest to get
"the time the virtual processor consumes running guest code,
and the time the associated logical processor spends running
hypervisor code on behalf of that guest."

Calculation of this time is performed by task_cputime_adjusted()
for vcpu task.

Necessary to support loading of winhv.sys in guest, which in turn is
required to support Windows VMBus.

Signed-off-by: Andrey Smetanin <asmetanin@virtuozzo.com>
Reviewed-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Paolo Bonzini <pbonzini@redhat.com>
CC: Gleb Natapov <gleb@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/asm/kvm_host.h    |  1 +
 arch/x86/include/uapi/asm/hyperv.h |  3 +++
 arch/x86/kvm/hyperv.c              | 21 +++++++++++++++++++--
 arch/x86/kvm/x86.c                 |  1 +
 kernel/sched/cputime.c             |  2 ++
 5 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 76a5b30979b3..d064cb2e19e8 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -374,6 +374,7 @@ struct kvm_mtrr {
 /* Hyper-V per vcpu emulation context */
 struct kvm_vcpu_hv {
 	u64 hv_vapic;
+	s64 runtime_offset;
 };
 
 struct kvm_vcpu_arch {
diff --git a/arch/x86/include/uapi/asm/hyperv.h b/arch/x86/include/uapi/asm/hyperv.h
index dab584bf7ddf..2677a0aac2cc 100644
--- a/arch/x86/include/uapi/asm/hyperv.h
+++ b/arch/x86/include/uapi/asm/hyperv.h
@@ -156,6 +156,9 @@
 /* MSR used to reset the guest OS. */
 #define HV_X64_MSR_RESET			0x40000003
 
+/* MSR used to provide vcpu runtime in 100ns units */
+#define HV_X64_MSR_VP_RUNTIME			0x40000010
+
 /* MSR used to read the per-partition time reference counter */
 #define HV_X64_MSR_TIME_REF_COUNT		0x40000020
 
diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
index 0ad11a232474..62cf8c915e95 100644
--- a/arch/x86/kvm/hyperv.c
+++ b/arch/x86/kvm/hyperv.c
@@ -178,7 +178,16 @@ static int kvm_hv_set_msr_pw(struct kvm_vcpu *vcpu, u32 msr, u64 data,
 	return 0;
 }
 
-static int kvm_hv_set_msr(struct kvm_vcpu *vcpu, u32 msr, u64 data)
+/* Calculate cpu time spent by current task in 100ns units */
+static u64 current_task_runtime_100ns(void)
+{
+	cputime_t utime, stime;
+
+	task_cputime_adjusted(current, &utime, &stime);
+	return div_u64(cputime_to_nsecs(utime + stime), 100);
+}
+
+static int kvm_hv_set_msr(struct kvm_vcpu *vcpu, u32 msr, u64 data, bool host)
 {
 	struct kvm_vcpu_hv *hv = &vcpu->arch.hyperv;
 
@@ -212,6 +221,11 @@ static int kvm_hv_set_msr(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 		return kvm_hv_vapic_msr_write(vcpu, APIC_ICR, data);
 	case HV_X64_MSR_TPR:
 		return kvm_hv_vapic_msr_write(vcpu, APIC_TASKPRI, data);
+	case HV_X64_MSR_VP_RUNTIME:
+		if (!host)
+			return 1;
+		hv->runtime_offset = data - current_task_runtime_100ns();
+		break;
 	default:
 		vcpu_unimpl(vcpu, "Hyper-V uhandled wrmsr: 0x%x data 0x%llx\n",
 			    msr, data);
@@ -287,6 +301,9 @@ static int kvm_hv_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 	case HV_X64_MSR_APIC_ASSIST_PAGE:
 		data = hv->hv_vapic;
 		break;
+	case HV_X64_MSR_VP_RUNTIME:
+		data = current_task_runtime_100ns() + hv->runtime_offset;
+		break;
 	default:
 		vcpu_unimpl(vcpu, "Hyper-V unhandled rdmsr: 0x%x\n", msr);
 		return 1;
@@ -305,7 +322,7 @@ int kvm_hv_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data, bool host)
 		mutex_unlock(&vcpu->kvm->lock);
 		return r;
 	} else
-		return kvm_hv_set_msr(vcpu, msr, data);
+		return kvm_hv_set_msr(vcpu, msr, data, host);
 }
 
 int kvm_hv_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5bc598de28a7..3e4d03220650 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -954,6 +954,7 @@ static u32 emulated_msrs[] = {
 	HV_X64_MSR_CRASH_P3, HV_X64_MSR_CRASH_P4, HV_X64_MSR_CRASH_CTL,
 	HV_X64_MSR_RESET,
 	HV_X64_MSR_VP_INDEX,
+	HV_X64_MSR_VP_RUNTIME,
 	HV_X64_MSR_APIC_ASSIST_PAGE, MSR_KVM_ASYNC_PF_EN, MSR_KVM_STEAL_TIME,
 	MSR_KVM_PV_EOI_EN,
 
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 8cbc3db671df..26a54461bf59 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -444,6 +444,7 @@ void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
 	*ut = p->utime;
 	*st = p->stime;
 }
+EXPORT_SYMBOL_GPL(task_cputime_adjusted);
 
 void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
 {
@@ -652,6 +653,7 @@ void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
 	task_cputime(p, &cputime.utime, &cputime.stime);
 	cputime_adjust(&cputime, &p->prev_cputime, ut, st);
 }
+EXPORT_SYMBOL_GPL(task_cputime_adjusted);
 
 void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
 {
-- 
1.8.3.1

[FYI PATCH 16/20] KVM: VMX: adjust interface to allocate/free_vpid

[Paolo Bonzini]

From: Wanpeng Li <wanpeng.li@hotmail.com>

Adjust allocate/free_vid so that they can be reused for the nested vpid.

Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/vmx.c | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index b2d619ab3d07..656074153f49 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -4155,29 +4155,28 @@ static int alloc_identity_pagetable(struct kvm *kvm)
 	return r;
 }
 
-static void allocate_vpid(struct vcpu_vmx *vmx)
+static int allocate_vpid(void)
 {
 	int vpid;
 
-	vmx->vpid = 0;
 	if (!enable_vpid)
-		return;
+		return 0;
 	spin_lock(&vmx_vpid_lock);
 	vpid = find_first_zero_bit(vmx_vpid_bitmap, VMX_NR_VPIDS);
-	if (vpid < VMX_NR_VPIDS) {
-		vmx->vpid = vpid;
+	if (vpid < VMX_NR_VPIDS)
 		__set_bit(vpid, vmx_vpid_bitmap);
-	}
+	else
+		vpid = 0;
 	spin_unlock(&vmx_vpid_lock);
+	return vpid;
 }
 
-static void free_vpid(struct vcpu_vmx *vmx)
+static void free_vpid(int vpid)
 {
-	if (!enable_vpid)
+	if (!enable_vpid || vpid == 0)
 		return;
 	spin_lock(&vmx_vpid_lock);
-	if (vmx->vpid != 0)
-		__clear_bit(vmx->vpid, vmx_vpid_bitmap);
+	__clear_bit(vpid, vmx_vpid_bitmap);
 	spin_unlock(&vmx_vpid_lock);
 }
 
@@ -8483,7 +8482,7 @@ static void vmx_free_vcpu(struct kvm_vcpu *vcpu)
 
 	if (enable_pml)
 		vmx_disable_pml(vmx);
-	free_vpid(vmx);
+	free_vpid(vmx->vpid);
 	leave_guest_mode(vcpu);
 	vmx_load_vmcs01(vcpu);
 	free_nested(vmx);
@@ -8502,7 +8501,7 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 	if (!vmx)
 		return ERR_PTR(-ENOMEM);
 
-	allocate_vpid(vmx);
+	vmx->vpid = allocate_vpid();
 
 	err = kvm_vcpu_init(&vmx->vcpu, kvm, id);
 	if (err)
@@ -8578,7 +8577,7 @@ free_msrs:
 uninit_vcpu:
 	kvm_vcpu_uninit(&vmx->vcpu);
 free_vcpu:
-	free_vpid(vmx);
+	free_vpid(vmx->vpid);
 	kmem_cache_free(kvm_vcpu_cache, vmx);
 	return ERR_PTR(err);
 }
-- 
1.8.3.1

[FYI PATCH 17/20] KVM: VMX: introduce __vmx_flush_tlb to handle specific vpid

[Paolo Bonzini]

From: Wanpeng Li <wanpeng.li@hotmail.com>

Introduce __vmx_flush_tlb() to handle specific vpid.

Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/vmx.c | 21 +++++++++++++--------
 1 file changed, 13 insertions(+), 8 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 656074153f49..e1cb4aa026a2 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -1337,13 +1337,13 @@ static void loaded_vmcs_clear(struct loaded_vmcs *loaded_vmcs)
 			 __loaded_vmcs_clear, loaded_vmcs, 1);
 }
 
-static inline void vpid_sync_vcpu_single(struct vcpu_vmx *vmx)
+static inline void vpid_sync_vcpu_single(int vpid)
 {
-	if (vmx->vpid == 0)
+	if (vpid == 0)
 		return;
 
 	if (cpu_has_vmx_invvpid_single())
-		__invvpid(VMX_VPID_EXTENT_SINGLE_CONTEXT, vmx->vpid, 0);
+		__invvpid(VMX_VPID_EXTENT_SINGLE_CONTEXT, vpid, 0);
 }
 
 static inline void vpid_sync_vcpu_global(void)
@@ -1352,10 +1352,10 @@ static inline void vpid_sync_vcpu_global(void)
 		__invvpid(VMX_VPID_EXTENT_ALL_CONTEXT, 0, 0);
 }
 
-static inline void vpid_sync_context(struct vcpu_vmx *vmx)
+static inline void vpid_sync_context(int vpid)
 {
 	if (cpu_has_vmx_invvpid_single())
-		vpid_sync_vcpu_single(vmx);
+		vpid_sync_vcpu_single(vpid);
 	else
 		vpid_sync_vcpu_global();
 }
@@ -3441,9 +3441,9 @@ static void exit_lmode(struct kvm_vcpu *vcpu)
 
 #endif
 
-static void vmx_flush_tlb(struct kvm_vcpu *vcpu)
+static inline void __vmx_flush_tlb(struct kvm_vcpu *vcpu, int vpid)
 {
-	vpid_sync_context(to_vmx(vcpu));
+	vpid_sync_context(vpid);
 	if (enable_ept) {
 		if (!VALID_PAGE(vcpu->arch.mmu.root_hpa))
 			return;
@@ -3451,6 +3451,11 @@ static void vmx_flush_tlb(struct kvm_vcpu *vcpu)
 	}
 }
 
+static void vmx_flush_tlb(struct kvm_vcpu *vcpu)
+{
+	__vmx_flush_tlb(vcpu, to_vmx(vcpu)->vpid);
+}
+
 static void vmx_decache_cr0_guest_bits(struct kvm_vcpu *vcpu)
 {
 	ulong cr0_guest_owned_bits = vcpu->arch.cr0_guest_owned_bits;
@@ -4784,7 +4789,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
 	vmx_fpu_activate(vcpu);
 	update_exception_bitmap(vcpu);
 
-	vpid_sync_context(vmx);
+	vpid_sync_context(vmx->vpid);
 }
 
 /*
-- 
1.8.3.1

[FYI PATCH 18/20] KVM: nVMX: emulate the INVVPID instruction

[Paolo Bonzini]

From: Wanpeng Li <wanpeng.li@hotmail.com>

Add the INVVPID instruction emulation.

Reviewed-by: Wincy Van <fanwenyi0529@gmail.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/asm/vmx.h |  1 +
 arch/x86/kvm/vmx.c         | 23 ++++++++++++++++++++++-
 2 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/arch/x86/include/asm/vmx.h b/arch/x86/include/asm/vmx.h
index 448b7ca61aee..53c9567a29d1 100644
--- a/arch/x86/include/asm/vmx.h
+++ b/arch/x86/include/asm/vmx.h
@@ -397,6 +397,7 @@ enum vmcs_field {
 #define IDENTITY_PAGETABLE_PRIVATE_MEMSLOT	(KVM_USER_MEM_SLOTS + 2)
 
 #define VMX_NR_VPIDS				(1 << 16)
+#define VMX_VPID_EXTENT_INDIVIDUAL_ADDR 	0
 #define VMX_VPID_EXTENT_SINGLE_CONTEXT		1
 #define VMX_VPID_EXTENT_ALL_CONTEXT		2
 
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index e1cb4aa026a2..bb7ae30caa13 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -7186,7 +7186,28 @@ static int handle_invept(struct kvm_vcpu *vcpu)
 
 static int handle_invvpid(struct kvm_vcpu *vcpu)
 {
-	kvm_queue_exception(vcpu, UD_VECTOR);
+	u32 vmx_instruction_info;
+	unsigned long type;
+
+	if (!nested_vmx_check_permission(vcpu))
+		return 1;
+
+	vmx_instruction_info = vmcs_read32(VMX_INSTRUCTION_INFO);
+	type = kvm_register_readl(vcpu, (vmx_instruction_info >> 28) & 0xf);
+
+	switch (type) {
+	case VMX_VPID_EXTENT_INDIVIDUAL_ADDR:
+	case VMX_VPID_EXTENT_SINGLE_CONTEXT:
+	case VMX_VPID_EXTENT_ALL_CONTEXT:
+		vmx_flush_tlb(vcpu);
+		nested_vmx_succeed(vcpu);
+		break;
+	default:
+		nested_vmx_failInvalid(vcpu);
+		break;
+	}
+
+	skip_emulated_instruction(vcpu);
 	return 1;
 }
 
-- 
1.8.3.1

[FYI PATCH 19/20] KVM: nVMX: nested VPID emulation

[Paolo Bonzini]

From: Wanpeng Li <wanpeng.li@hotmail.com>

VPID is used to tag address space and avoid a TLB flush. Currently L0 use
the same VPID to run L1 and all its guests. KVM flushes VPID when switching
between L1 and L2.

This patch advertises VPID to the L1 hypervisor, then address space of L1
and L2 can be separately treated and avoid TLB flush when swithing between
L1 and L2. For each nested vmentry, if vpid12 is changed, reuse shadow vpid
w/ an invvpid.

Performance:

run lmbench on L2 w/ 3.5 kernel.

Context switching - times in microseconds - smaller is better
-------------------------------------------------------------------------
Host                 OS  2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
                         ctxsw  ctxsw  ctxsw ctxsw  ctxsw   ctxsw   ctxsw
--------- ------------- ------ ------ ------ ------ ------ ------- -------
kernel    Linux 3.5.0-1 1.2200 1.3700 1.4500 4.7800 2.3300 5.60000 2.88000  nested VPID
kernel    Linux 3.5.0-1 1.2600 1.4300 1.5600   12.7   12.9 3.49000 7.46000  vanilla

Reviewed-by: Jan Kiszka <jan.kiszka@siemens.com>
Suggested-by: Wincy Van <fanwenyi0529@gmail.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
[Handle the case where vpid02 cannot be allocated. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/vmx.c | 39 ++++++++++++++++++++++++++++++++-------
 1 file changed, 32 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index bb7ae30caa13..75f3ee01f59b 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -424,6 +424,9 @@ struct nested_vmx {
 	/* to migrate it to L2 if VM_ENTRY_LOAD_DEBUG_CONTROLS is off */
 	u64 vmcs01_debugctl;
 
+	u16 vpid02;
+	u16 last_vpid;
+
 	u32 nested_vmx_procbased_ctls_low;
 	u32 nested_vmx_procbased_ctls_high;
 	u32 nested_vmx_true_procbased_ctls_low;
@@ -1157,6 +1160,11 @@ static inline bool nested_cpu_has_virt_x2apic_mode(struct vmcs12 *vmcs12)
 	return nested_cpu_has2(vmcs12, SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE);
 }
 
+static inline bool nested_cpu_has_vpid(struct vmcs12 *vmcs12)
+{
+	return nested_cpu_has2(vmcs12, SECONDARY_EXEC_ENABLE_VPID);
+}
+
 static inline bool nested_cpu_has_apic_reg_virt(struct vmcs12 *vmcs12)
 {
 	return nested_cpu_has2(vmcs12, SECONDARY_EXEC_APIC_REGISTER_VIRT);
@@ -2471,6 +2479,7 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx)
 		SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES |
 		SECONDARY_EXEC_RDTSCP |
 		SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE |
+		SECONDARY_EXEC_ENABLE_VPID |
 		SECONDARY_EXEC_APIC_REGISTER_VIRT |
 		SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY |
 		SECONDARY_EXEC_WBINVD_EXITING |
@@ -6670,6 +6679,7 @@ static void free_nested(struct vcpu_vmx *vmx)
 		return;
 
 	vmx->nested.vmxon = false;
+	free_vpid(vmx->nested.vpid02);
 	nested_release_vmcs12(vmx);
 	if (enable_shadow_vmcs)
 		free_vmcs(vmx->nested.current_shadow_vmcs);
@@ -7199,7 +7209,7 @@ static int handle_invvpid(struct kvm_vcpu *vcpu)
 	case VMX_VPID_EXTENT_INDIVIDUAL_ADDR:
 	case VMX_VPID_EXTENT_SINGLE_CONTEXT:
 	case VMX_VPID_EXTENT_ALL_CONTEXT:
-		vmx_flush_tlb(vcpu);
+		__vmx_flush_tlb(vcpu, to_vmx(vcpu)->nested.vpid02);
 		nested_vmx_succeed(vcpu);
 		break;
 	default:
@@ -8575,8 +8585,10 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 			goto free_vmcs;
 	}
 
-	if (nested)
+	if (nested) {
 		nested_vmx_setup_ctls_msrs(vmx);
+		vmx->nested.vpid02 = allocate_vpid();
+	}
 
 	vmx->nested.posted_intr_nv = -1;
 	vmx->nested.current_vmptr = -1ull;
@@ -8597,6 +8609,7 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 	return &vmx->vcpu;
 
 free_vmcs:
+	free_vpid(vmx->nested.vpid02);
 	free_loaded_vmcs(vmx->loaded_vmcs);
 free_msrs:
 	kfree(vmx->guest_msrs);
@@ -9458,12 +9471,24 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
 
 	if (enable_vpid) {
 		/*
-		 * Trivially support vpid by letting L2s share their parent
-		 * L1's vpid. TODO: move to a more elaborate solution, giving
-		 * each L2 its own vpid and exposing the vpid feature to L1.
+		 * There is no direct mapping between vpid02 and vpid12, the
+		 * vpid02 is per-vCPU for L0 and reused while the value of
+		 * vpid12 is changed w/ one invvpid during nested vmentry.
+		 * The vpid12 is allocated by L1 for L2, so it will not
+		 * influence global bitmap(for vpid01 and vpid02 allocation)
+		 * even if spawn a lot of nested vCPUs.
 		 */
-		vmcs_write16(VIRTUAL_PROCESSOR_ID, vmx->vpid);
-		vmx_flush_tlb(vcpu);
+		if (nested_cpu_has_vpid(vmcs12) && vmx->nested.vpid02) {
+			vmcs_write16(VIRTUAL_PROCESSOR_ID, vmx->nested.vpid02);
+			if (vmcs12->virtual_processor_id != vmx->nested.last_vpid) {
+				vmx->nested.last_vpid = vmcs12->virtual_processor_id;
+				__vmx_flush_tlb(vcpu, to_vmx(vcpu)->nested.vpid02);
+			}
+		} else {
+			vmcs_write16(VIRTUAL_PROCESSOR_ID, vmx->vpid);
+			vmx_flush_tlb(vcpu);
+		}
+
 	}
 
 	if (nested_cpu_has_ept(vmcs12)) {
-- 
1.8.3.1

[FYI PATCH 20/20] KVM: vmx: disable posted interrupts if no local APIC

[Paolo Bonzini]

Uniprocessor 32-bit randconfigs can disable the local APIC, and posted
interrupts require reserving a vector on the LAPIC, so they are
incompatible.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/vmx.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 75f3ee01f59b..353b91b744d7 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -986,7 +986,8 @@ static inline bool cpu_has_vmx_virtual_intr_delivery(void)
 
 static inline bool cpu_has_vmx_posted_intr(void)
 {
-	return vmcs_config.pin_based_exec_ctrl & PIN_BASED_POSTED_INTR;
+	return IS_ENABLED(CONFIG_X86_LOCAL_APIC) &&
+		vmcs_config.pin_based_exec_ctrl & PIN_BASED_POSTED_INTR;
 }
 
 static inline bool cpu_has_vmx_apicv(void)
-- 
1.8.3.1