diff mbox series

[2/3] KVM: X86: Implement PV sched yield hypercall

Message ID 1558953255-9432-3-git-send-email-wanpengli@tencent.com (mailing list archive)
State New, archived
Headers show
Series KVM: Yield to IPI target if necessary | expand

Commit Message

Wanpeng Li May 27, 2019, 10:34 a.m. UTC 
From: Wanpeng Li <wanpengli@tencent.com>

The target vCPUs are in runnable state after vcpu_kick and suitable 
as a yield target. This patch implements the sched yield hypercall.

17% performace increase of ebizzy benchmark can be observed in an 
over-subscribe environment. (w/ kvm-pv-tlb disabled, testing TLB flush 
call-function IPI-many since call-function is not easy to be trigged 
by userspace workload).

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
 arch/x86/kvm/x86.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

Comments

Paolo Bonzini May 27, 2019, 11:53 a.m. UTC | #1 
On 27/05/19 12:34, Wanpeng Li wrote:
> +	rcu_read_lock();
> +	map = rcu_dereference(kvm->arch.apic_map);
> +	target = map->phys_map[dest_id]->vcpu;
> +	rcu_read_unlock();
> +
> +	kvm_vcpu_yield_to(target);

This needs to check that map->phys_map[dest_id] is not NULL.

Paolo
Wanpeng Li May 28, 2019, 12:54 a.m. UTC | #2 
On Mon, 27 May 2019 at 19:54, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> On 27/05/19 12:34, Wanpeng Li wrote:
> > +     rcu_read_lock();
> > +     map = rcu_dereference(kvm->arch.apic_map);
> > +     target = map->phys_map[dest_id]->vcpu;
> > +     rcu_read_unlock();
> > +
> > +     kvm_vcpu_yield_to(target);
>
> This needs to check that map->phys_map[dest_id] is not NULL.

Fixed in v2, thanks for the quick review. :)

Regards,
Wanpeng Li
diff mbox series

Patch

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index e7e57de..45403a6 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7172,6 +7172,19 @@  void kvm_vcpu_deactivate_apicv(struct kvm_vcpu *vcpu)
 	kvm_x86_ops->refresh_apicv_exec_ctrl(vcpu);
 }
 
+void kvm_sched_yield(struct kvm *kvm, u64 dest_id)
+{
+	struct kvm_vcpu *target;
+	struct kvm_apic_map *map;
+
+	rcu_read_lock();
+	map = rcu_dereference(kvm->arch.apic_map);
+	target = map->phys_map[dest_id]->vcpu;
+	rcu_read_unlock();
+
+	kvm_vcpu_yield_to(target);
+}
+
 int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
 {
 	unsigned long nr, a0, a1, a2, a3, ret;
@@ -7218,6 +7231,10 @@  int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
 	case KVM_HC_SEND_IPI:
 		ret = kvm_pv_send_ipi(vcpu->kvm, a0, a1, a2, a3, op_64_bit);
 		break;
+	case KVM_HC_SCHED_YIELD:
+		kvm_sched_yield(vcpu->kvm, a0);
+		ret = 0;
+		break;
 	default:
 		ret = -KVM_ENOSYS;
 		break;