Re: CFS group scheduler fairness broken starting from 2.6.29-rc1
From: Ken Chen
Date: Thu Jul 23 2009 - 18:17:40 EST
On Thu, Jul 23, 2009 at 12:57 AM, Bharata B
Rao<bharata@xxxxxxxxxxxxxxxxxx> wrote:
> Hi,
>
> Group scheduler fainess is broken since 2.6.29-rc1. git bisect led me
> to this commit:
>
> commit ec4e0e2fe018992d980910db901637c814575914
> Author: Ken Chen <kenchen@xxxxxxxxxx>
> Date: Tue Nov 18 22:41:57 2008 -0800
>
> sched: fix inconsistency when redistribute per-cpu tg->cfs_rq shares
>
> Impact: make load-balancing more consistent
> ....
>
> ======================================================================
> % CPU time division b/n groups
> Group 2.6.29-rc1 2.6.29-rc1 w/o the above patch
> ======================================================================
> a with 8 tasks 44 31
> b with 5 tasks 32 34
> c with 3 tasks 22 34
> ======================================================================
> All groups had equal shares.
What value did you use for each task_group's share? For very large
value of tg->shares, it could be that all of the boost went to one CPU
and subsequently causes load-balancer to shuffle tasks around. Do you
see any unexpected task migration?
- Ken
¢éì¹»®&Þ~º&¶¬?+-±éݶ¥?w®?Ë?±Êâméb?ìdz¹Þ?)í?æèw*jg¬±¨¶????Ý¢j/?êäz¹Þ??à2?Þ?¨èÚ&¢)ß¡«a¶Úþø®G«?éh®æj:+v?¨?wè?Ù¥>W?±êÞiÛaxPjØm¶?ÿÃ-»+?ùd?_