Messages in this thread | | | Subject | Re: [RFC PATCH v2] sched: Limit idle_balance() | From | Jason Low <> | Date | Mon, 22 Jul 2013 10:42:23 -0700 |
| |
On Sun, 2013-07-21 at 23:02 +0530, Preeti U Murthy wrote: > Hi Json, > > With V2 of your patch here are the results for the ebizzy run on > 3.11-rc1 + patch on a 1 socket, 16 core powerpc machine. Each ebizzy > run was for 30 seconds. > > Number_of_threads %improvement_with_patch > 4 8.63 > 8 1.29 > 12 9.98 > 16 20.46 > > Let me know if you want me to profile any of these runs for specific > statistics.
Hi Preeti,
Other numbers that would be interesting to know are the % improvements when N = 20. That is when I'm getting the bigger performance boosts on some of the AIM7 workloads.
Would you like to re-run ebizzy with v2 of my patch except replacing this line of code: if (idle_duration / 10 < (sd->avg_idle_balance_cost + cost)) with this line of code: if (idle_duration / 20 < (sd->avg_idle_balance_cost + cost))
Thanks a lot, Jason
| |