[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 12 Dec 2019 16:06:38 +0100
From: Dietmar Eggemann <dietmar.eggemann@....com>
To: Valentin Schneider <valentin.schneider@....com>,
linux-kernel@...r.kernel.org
Cc: peterz@...radead.org, mingo@...nel.org, vincent.guittot@...aro.org,
patrick.bellasi@...bug.net, qperret@...gle.com,
qais.yousef@....com, morten.rasmussen@....com
Subject: Re: [PATCH v3 0/5] sched/fair: Task placement biasing using uclamp
On 11/12/2019 12:38, Valentin Schneider wrote:
> Hi,
>
> While uclamp restrictions currently only impact schedutil's frequency
> selection, it would make sense to also let them impact CPU selection in
> asymmetric topologies. This would let us steer specific tasks towards
> certain CPU capacities regardless of their actual utilization - I give a
> few examples in patch 4.
>
> The first three patches are mainly cleanups, the meat of the thing is
> in patches 4 and 5.
>
> Note that this is in the same spirit as what Patrick had proposed for EAS
> on Android [1]
>
> [1]: https://android.googlesource.com/kernel/common/+/b61876ed122f816660fe49e0de1b7ee4891deaa2%5E%21
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@....com>
Tested-By: Dietmar Eggemann <dietmar.eggemann@....com>
Tested on Juno-r0 (Arm64) cpumask [0x3f] w/ big [0x06], LITTLE [0x39]
[orig cpu capacity big,LITTLE: 1024,446] and rt-app
4 periodic tasks runtime/period [800/16000], per task uclamp_min/max
[600,1024]
w/o uclamp: EAS puts the tasks on LITTLE CPUs [0x39]
w/ uclamp: EAS puts the tasks on big CPUs [0x06]
Powered by blists - more mailing lists