[v0,11/15] KVM: x86/xen: handle PV timers oneshot mode

Commit Message

David Woodhouse Feb. 10, 2022, 12:27 a.m. UTC

From: Joao Martins <[e-mail]>

If the guest has offloaded the timer virq, handle the following
hypercalls for programming the timer:

    VCPUOP_set_singleshot_timer
    VCPUOP_stop_singleshot_timer
    set_timer_op(timestamp_ns)

The event channel corresponding to the timer virq is then used to inject
events once timer deadlines are met. For now we back the PV timer with
hrtimer.

Signed-off-by: Joao Martins <[e-mail]>
Signed-off-by: David Woodhouse <[e-mail]>
---
 arch/x86/include/asm/kvm_host.h |   3 +
 arch/x86/kvm/irq.c              |  11 ++-
 arch/x86/kvm/x86.c              |   3 +
 arch/x86/kvm/xen.c              | 168 ++++++++++++++++++++++++++++++++
 arch/x86/kvm/xen.h              |  30 ++++++
 include/uapi/linux/kvm.h        |   6 ++
 6 files changed, 219 insertions(+), 2 deletions(-)

Patch

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 3c58d0bf5f9b..ddfdc8cc8e60 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -609,6 +609,9 @@  struct kvm_vcpu_xen {
 	u64 runstate_times[4];
 	unsigned long evtchn_pending_sel;
 	u32 vcpu_id; /* The Xen / ACPI vCPU ID */
+	u32 timer_virq;
+	atomic_t timer_pending;
+	struct hrtimer timer;
 };
 
 struct kvm_vcpu_arch {
diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c
index 172b05343cfd..af2d26fc5458 100644
--- a/arch/x86/kvm/irq.c
+++ b/arch/x86/kvm/irq.c
@@ -22,10 +22,14 @@ 
  */
 int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
 {
+	int r = 0;
+
 	if (lapic_in_kernel(vcpu))
-		return apic_has_pending_timer(vcpu);
+		r = apic_has_pending_timer(vcpu);
+	if (kvm_xen_timer_enabled(vcpu))
+		r += kvm_xen_has_pending_timer(vcpu);
 
-	return 0;
+	return r;
 }
 EXPORT_SYMBOL(kvm_cpu_has_pending_timer);
 
@@ -143,6 +147,8 @@  void kvm_inject_pending_timer_irqs(struct kvm_vcpu *vcpu)
 {
 	if (lapic_in_kernel(vcpu))
 		kvm_inject_apic_timer_irqs(vcpu);
+	if (kvm_xen_timer_enabled(vcpu))
+		kvm_xen_inject_timer_irqs(vcpu);
 }
 EXPORT_SYMBOL_GPL(kvm_inject_pending_timer_irqs);
 
@@ -150,6 +156,7 @@  void __kvm_migrate_timers(struct kvm_vcpu *vcpu)
 {
 	__kvm_migrate_apic_timer(vcpu);
 	__kvm_migrate_pit_timer(vcpu);
+	__kvm_migrate_xen_timer(vcpu);
 	static_call_cond(kvm_x86_migrate_timers)(vcpu);
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 2143c2652f8f..7240d791e4ab 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -12063,6 +12063,9 @@  static inline bool kvm_vcpu_has_events(struct kvm_vcpu *vcpu)
 	    kvm_x86_ops.nested_ops->hv_timer_pending(vcpu))
 		return true;
 
+	if (kvm_xen_has_pending_timer(vcpu))
+		return true;
+
 	return false;
 }
 
diff --git a/arch/x86/kvm/xen.c b/arch/x86/kvm/xen.c
index 8ee4bc648bcb..c042c7d6ee02 100644
--- a/arch/x86/kvm/xen.c
+++ b/arch/x86/kvm/xen.c
@@ -23,6 +23,7 @@ 
 
 #include "trace.h"
 
+static int kvm_xen_set_evtchn(struct kvm_xen_evtchn *xe, struct kvm *kvm);
 static int kvm_xen_setattr_evtchn(struct kvm *kvm, struct kvm_xen_hvm_attr *data);
 static bool kvm_xen_hcall_evtchn_send(struct kvm_vcpu *vcpu, u64 param, u64 *r);
 
@@ -108,6 +109,80 @@  static int kvm_xen_shared_info_init(struct kvm *kvm, gfn_t gfn)
 	return ret;
 }
 
+void kvm_xen_inject_timer_irqs(struct kvm_vcpu *vcpu)
+{
+	if (atomic_read(&vcpu->arch.xen.timer_pending) > 0) {
+		struct kvm_xen_evtchn e;
+
+		e.vcpu_id = vcpu->vcpu_id;
+		e.vcpu_idx = vcpu->vcpu_idx;
+		e.port = vcpu->arch.xen.timer_virq;
+		e.priority = KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL;
+
+		kvm_xen_set_evtchn(&e, vcpu->kvm);
+		atomic_set(&vcpu->arch.xen.timer_pending, 0);
+	}
+}
+
+static enum hrtimer_restart xen_timer_callback(struct hrtimer *timer)
+{
+	struct kvm_vcpu *vcpu = container_of(timer, struct kvm_vcpu,
+					     arch.xen.timer);
+	struct kvm_xen_evtchn e;
+
+	if (atomic_read(&vcpu->arch.xen.timer_pending))
+		return HRTIMER_NORESTART;
+
+	e.vcpu_id = vcpu->vcpu_id;
+	e.vcpu_idx = vcpu->vcpu_idx;
+	e.port = vcpu->arch.xen.timer_virq;
+	e.priority = KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL;
+
+	if (kvm_xen_set_evtchn_fast(&e, vcpu->kvm) != -EWOULDBLOCK)
+		return HRTIMER_NORESTART;
+
+	atomic_inc(&vcpu->arch.xen.timer_pending);
+	kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
+	kvm_vcpu_kick(vcpu);
+
+	return HRTIMER_NORESTART;
+}
+
+void __kvm_migrate_xen_timer(struct kvm_vcpu *vcpu)
+{
+	struct hrtimer *timer;
+
+	if (!kvm_xen_timer_enabled(vcpu))
+		return;
+
+	timer = &vcpu->arch.xen.timer;
+	if (hrtimer_cancel(timer))
+		hrtimer_start_expires(timer, HRTIMER_MODE_ABS_PINNED);
+}
+
+static void kvm_xen_start_timer(struct kvm_vcpu *vcpu, u64 delta_ns)
+{
+	ktime_t ktime_now;
+
+	atomic_set(&vcpu->arch.xen.timer_pending, 0);
+	ktime_now = ktime_get();
+	hrtimer_start(&vcpu->arch.xen.timer,
+		      ktime_add_ns(ktime_now, delta_ns),
+		      HRTIMER_MODE_ABS_PINNED);
+}
+
+static void kvm_xen_stop_timer(struct kvm_vcpu *vcpu)
+{
+	hrtimer_cancel(&vcpu->arch.xen.timer);
+}
+
+void kvm_xen_init_timer(struct kvm_vcpu *vcpu)
+{
+	hrtimer_init(&vcpu->arch.xen.timer, CLOCK_MONOTONIC,
+		     HRTIMER_MODE_ABS_PINNED);
+	vcpu->arch.xen.timer.function = xen_timer_callback;
+}
+
 static void kvm_xen_update_runstate(struct kvm_vcpu *v, int state)
 {
 	struct kvm_vcpu_xen *vx = &v->arch.xen;
@@ -614,6 +689,21 @@  int kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)
 		}
 		break;
 
+	case KVM_XEN_VCPU_ATTR_TYPE_TIMER:
+		if (data->u.timer.port) {
+			if (data->u.timer.priority != KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL) {
+				r = -EINVAL;
+				break;
+			}
+			kvm_xen_init_timer(vcpu);
+		} else if (kvm_xen_timer_enabled(vcpu)) {
+			kvm_xen_stop_timer(vcpu);
+		}
+		vcpu->arch.xen.timer_virq = data->u.timer.port;
+		// XXX Start timer if it's given
+		r = 0;
+		break;
+
 	default:
 		break;
 	}
@@ -694,6 +784,13 @@  int kvm_xen_vcpu_get_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)
 		r = 0;
 		break;
 
+	case KVM_XEN_VCPU_ATTR_TYPE_TIMER:
+		data->u.timer.port = vcpu->arch.xen.timer_virq;
+		data->u.timer.priority = KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL;
+		data->u.timer.expires_ns = 0; // XXX
+		r = 0;
+		break;
+
 	default:
 		break;
 	}
@@ -855,6 +952,67 @@  static bool kvm_xen_hcall_sched_op(struct kvm_vcpu *vcpu, int cmd, u64 param, u6
 	return false;
 }
 
+static bool kvm_xen_hcall_vcpu_op(struct kvm_vcpu *vcpu, int cmd, int vcpu_id,
+				  u64 param, u64 *r)
+{
+	struct vcpu_set_singleshot_timer oneshot;
+	struct kvm_vcpu *target_vcpu;
+	long delta;
+	gpa_t gpa;
+	int idx;
+
+	target_vcpu = kvm_xen_vcpu_by_id(vcpu->kvm, vcpu_id, vcpu);
+	if (!target_vcpu || !kvm_xen_timer_enabled(target_vcpu))
+		return false;
+
+	switch (cmd) {
+	case VCPUOP_set_singleshot_timer:
+		idx = srcu_read_lock(&vcpu->kvm->srcu);
+		gpa = kvm_mmu_gva_to_gpa_system(vcpu, param, NULL);
+		srcu_read_unlock(&vcpu->kvm->srcu, idx);
+
+		if (!gpa || kvm_vcpu_read_guest(vcpu, gpa, &oneshot,
+						sizeof(oneshot))) {
+			*r = -EFAULT;
+			return true;
+		}
+
+		delta = oneshot.timeout_abs_ns - get_kvmclock_ns(vcpu->kvm);
+		kvm_xen_start_timer(target_vcpu, delta);
+		*r = 0;
+		return true;
+
+	case VCPUOP_stop_singleshot_timer:
+		kvm_xen_stop_timer(target_vcpu);
+		*r = 0;
+		return true;
+	}
+
+	return false;
+}
+
+static bool kvm_xen_hcall_set_timer_op(struct kvm_vcpu *vcpu, uint64_t timeout,
+				       u64 *r)
+{
+	ktime_t ktime_now = ktime_get();
+	long delta = timeout - get_kvmclock_ns(vcpu->kvm);
+
+	if (!kvm_xen_timer_enabled(vcpu))
+		return false;
+
+	if (timeout == 0) {
+		kvm_xen_stop_timer(vcpu);
+	} else if (unlikely(timeout < ktime_now) ||
+		   ((uint32_t) (delta >> 50) != 0)) {
+		kvm_xen_start_timer(vcpu, 50000000);
+	} else {
+		kvm_xen_start_timer(vcpu, delta);
+	}
+
+	*r = 0;
+	return true;
+}
+
 int kvm_xen_hypercall(struct kvm_vcpu *vcpu)
 {
 	bool longmode;
@@ -898,6 +1056,13 @@  int kvm_xen_hypercall(struct kvm_vcpu *vcpu)
 	case __HYPERVISOR_sched_op:
 		handled = kvm_xen_hcall_sched_op(vcpu, params[0], params[1], &r);
 		break;
+	case __HYPERVISOR_vcpu_op:
+		handled = kvm_xen_hcall_vcpu_op(vcpu, params[0], params[1],
+						params[2], &r);
+		break;
+	case __HYPERVISOR_set_timer_op:
+		handled = kvm_xen_hcall_set_timer_op(vcpu, params[0], &r);
+		break;
 	default:
 		break;
 	}
@@ -1405,6 +1570,9 @@  void kvm_xen_init_vcpu(struct kvm_vcpu *vcpu)
 
 void kvm_xen_destroy_vcpu(struct kvm_vcpu *vcpu)
 {
+	if (kvm_xen_timer_enabled(vcpu))
+		kvm_xen_stop_timer(vcpu);
+
 	kvm_gfn_to_pfn_cache_destroy(vcpu->kvm,
 				     &vcpu->arch.xen.runstate_cache);
 	kvm_gfn_to_pfn_cache_destroy(vcpu->kvm,
diff --git a/arch/x86/kvm/xen.h b/arch/x86/kvm/xen.h
index 54d587aae85b..616fe751c8fc 100644
--- a/arch/x86/kvm/xen.h
+++ b/arch/x86/kvm/xen.h
@@ -62,6 +62,21 @@  static inline bool kvm_xen_has_pending_events(struct kvm_vcpu *vcpu)
 		vcpu->arch.xen.evtchn_pending_sel;
 }
 
+static inline bool kvm_xen_timer_enabled(struct kvm_vcpu *vcpu)
+{
+	return !!vcpu->arch.xen.timer_virq;
+}
+
+static inline int kvm_xen_has_pending_timer(struct kvm_vcpu *vcpu)
+{
+	if (kvm_xen_hypercall_enabled(vcpu->kvm) && kvm_xen_timer_enabled(vcpu))
+		return atomic_read(&vcpu->arch.xen.timer_pending);
+
+	return 0;
+}
+
+void __kvm_migrate_xen_timer(struct kvm_vcpu *vcpu);
+void kvm_xen_inject_timer_irqs(struct kvm_vcpu *vcpu);
 #else
 static inline int kvm_xen_write_hypercall_page(struct kvm_vcpu *vcpu, u64 data)
 {
@@ -104,6 +119,21 @@  static inline void kvm_xen_inject_pending_events(struct kvm_vcpu *vcpu)
 }
 
 static inline bool kvm_xen_has_pending_events(struct kvm_vcpu *vcpu)
+
+static inline void __kvm_migrate_xen_timer(struct kvm_vcpu *vcpu)
+{
+}
+
+static inline int kvm_xen_has_pending_timer(struct kvm_vcpu *vcpu)
+{
+	return 0;
+}
+
+static inline void kvm_xen_inject_timer_irqs(struct kvm_vcpu *vcpu)
+{
+}
+
+static inline bool kvm_xen_timer_enabled(struct kvm_vcpu *vcpu)
 {
 	return false;
 }
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index 284645e1a872..f485b536112f 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -1740,6 +1740,11 @@  struct kvm_xen_vcpu_attr {
 			__u64 time_offline;
 		} runstate;
 		__u32 vcpu_id;
+		struct {
+			__u32 port;
+			__u32 priority;
+			__u64 expires_ns;
+		} timer;
 	} u;
 };
 
@@ -1752,6 +1757,7 @@  struct kvm_xen_vcpu_attr {
 #define KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST	0x5
 /* Available with KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_EVTCHN_SEND */
 #define KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID		0x6
+#define KVM_XEN_VCPU_ATTR_TYPE_TIMER		0x7
 
 /* Secure Encrypted Virtualization command */
 enum sev_cmd_id {