2 ms·
Can't find any numbers in the linked thread with the patches. Surely some preliminary benchmarking must have been performed that could tell us something about t
by simfoo 2y ago
Can't find any numbers in the linked thread with the patches. Surely some preliminary benchmarking must have been performed that could tell us something about the real world potential of the change?
- biorach 2y agoFrom the article, second last paragraph: > There is also, of course, the need for extensive performance testing; Mike Galbraith has made an early start on that work, showing that throughput with lazy preemption falls just short of that with PREEMPT_VOLUNTARY.
- spockz 2y agoHow would you benchmark something like this? Run multiple processes concurrently and then sort by total run time? Or measure individual process wait time?
- hifromwork 2y agoI guess both make sense, and a lot of other things (synthetical benchmarks, microbenchmarks, real-world benchmarks, best/average/worst case latency comparison, best/average/worst case throughput comparison...)