Re: [PATCH v4 2/3] sched/fair: Take thermal pressure into account while estimating energy
From: Lukasz Luba <lukasz.luba@arm.com>
Date: 2021-06-16 20:22:53
Also in:
lkml
On 6/16/21 8:25 PM, Vincent Guittot wrote:
On Wed, 16 Jun 2021 at 19:24, Dietmar Eggemann [off-list ref] wrote:quoted
On 15/06/2021 18:09, Lukasz Luba wrote:quoted
On 6/15/21 4:31 PM, Dietmar Eggemann wrote:quoted
On 14/06/2021 21:11, Lukasz Luba wrote:[...]quoted
quoted
It's important to highlight that this will only fix this issue between schedutil and EAS when it's due to `thermal pressure` (today only via CPU cooling). There are other places which could restrict policy->max via freq_qos_update_request() and EAS will be unaware of it.True, but for this I have some other plans.As long as people are aware of the fact that this was developed to be beneficial for `EAS - IPA` integration, I'm fine with this.I don't think it's only for EAS - IPA. Thermal_pressure can be used by HW throttling like here: https://lkml.org/lkml/2021/6/8/1791 EAS is involved but not IPA
Thank you Vincent for pointing to Thara's patches. Indeed, this is a good example. We will have to provide similar for our SCMI perf notifications - these are the plans that I've mentioned. In both new examples, the IPA (or other governors) won't be even involved.