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. Current Hyper-V TLFS (v5.0b) claims that HvCallSendSyntheticClusterIpi hypercall can't be 'fast' (passing parameters through registers) but apparently this is not true, Windows always uses it as 'fast' so we need to support that.
Showing the changes to include/uapi/linux (1 of 5 files). Show all
Unified · Split
| @@ -955,6 +955,7 @@ struct kvm_ppc_resize_hpt { · 954 unmodified lines | ||
| 955 | 955 | #define KVM_CAP_ARM_INJECT_SERROR_ESR 158 |
| 956 | 956 | #define KVM_CAP_MSR_PLATFORM_INFO 159 |
| 957 | 957 | #define KVM_CAP_PPC_NESTED_HV 160 |
| 958 | +#define KVM_CAP_HYPERV_SEND_IPI 161 | |
| 958 | 959 | |
| 959 | 960 | #ifdef KVM_CAP_IRQ_ROUTING |
| 960 | 961 | |