This feature introduces dedicated vCPUs(vCPU pinning, and there is no vCPU over-commitment) hint KVM_HINTS_DEDICATED, it has two users now:
Waiman Long mentioned that:
Generally speaking, unfair lock performs well for VMs with a small number of vCPUs. Native qspinlock may perform better than pvqspinlock if there is vCPU pinning and there is no vCPU over-commitment.
vCPUs are very unlikely to get preempted when they are the only task running on a CPU. PV TLB flush is slower than the native flush in that case.