Optimize nv_kthread_q batching and reduce per-item wakeup overhead#1050

arch-fan wants to merge 1 commit intoNVIDIA:mainNVIDIA/open-gpu-kernel-modules:mainfrom

arch-fan:perf/nv-kthread-q-batchingarch-fan/open-gpu-kernel-modules:perf/nv-kthread-q-batchingCopy head branch name to clipboard