Vitaly Kuznetsov <vkuznets@xxxxxxxxxx> writes: > Changes since v3 [Roman Kagan]: > - Don't use vcpu_id in kvm_make_vcpus_request_mask to avoid dealing with > sparse CPU masks. > - Create get_vcpu_idx_by_vpidx() and use it in get_vcpu_by_vpidx() and > kvm_hv_flush_tlb(). > > Using hypercall for sending IPIs is faster because this allows to specify > any number of vCPUs (even > 64 with sparse CPU set), the whole procedure > will take only one VMEXIT. > > Same as PV TLB flush, this allows Windows guests having > 64 vCPUs to boot > on KVM when Hyper-V extensions are enabled. Ping) -- Vitaly
