lkml.org 
[lkml]   [2023]   [Jul]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [Patch v3 4/6] sched/fair: Consider the idle state of the whole core for load balance
From
Date
On Fri, 2023-07-14 at 18:32 +0530, Shrikanth Hegde wrote:
>
>
>
> Tried on a symmetric system with all cores having SMT=4 as well. There was reduction in migrations here as well.
> Didnt observe any major regressions when microbenchmarks run alone. Such as hackbench, stress-ng.
>
> So. Here is tested-by.
> Tested-by: Shrikanth Hegde <sshegde@linux.vnet.ibm.com>

Thanks for testing.

>
>
> > +
> > /* Are we the first CPU of this group ? */
> > return group_balance_cpu(sg) == env->dst_cpu;
> > }
>
> One doubt though, Here a fully idle core would be chosen instead of first idle cpu in the
> group (if there is one). Since coming out of idle of SMT is faster compared to a fully idle core,
> would latency increase? Or that concerns mainly wakeup path?

Yeah, I think that concern is for the wakeup path and not for the balance path.

Tim

\
 
 \ /
  Last update: 2023-07-16 20:52    [W:6.140 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site