Thread (27 messages) 27 messages, 3 authors, 2021-11-09

Re: [PATCH v3 0/5] Refactor thermal pressure update to avoid code duplication

From: Lukasz Luba <lukasz.luba@arm.com>
Date: 2021-11-09 08:29:31
Also in: linux-arm-msm, linux-pm, lkml

Hi Steev,

That's interesting what you've done with Rockchip RK3399.
I would like to reproduce your experiment on my RockPI 4B v1.3.
Could you tell me how you to add this boost frequency that you have
mentioned in some previous emails?

I want to have similar setup to yours and I'll check all the subsystems
involved in the decision making process for triggering this boost freq.

On 11/8/21 11:21 PM, Steev Klimaszewski wrote:
Hi Thara,
quoted
Hi Steev,

IIUC, PineBook Pro has Rockchip RK3399 which has 2 Cortex A-72 and 4 
Cortex A-52 where as C630 has Qualcomm sdm845 which has 4 Cortex A-75 
and 4 Cortex A-55. Task placements and subsequently cpu load will be 
different for both the platforms. With the same workload, I will 
expect Rockchip to system to be more loaded than sdm845. Having said 
that, what cpu-freq governor are you using on both the systems.
I'm using sched-util on both of the systems.

I've tried a number of different ways of forcing builds only on the A-75 
cores, and I simply cannot get the load to be "enough" to kick in the 
boost frequency.

An example being

git clone https://github.com/zellij-org/zellij.git

cd zellij

taskset --cpu-list 4-7 cargo build --release

git clean -fdx

taskset --cpu-list 6-7 cargo build --release
Thanks for the pointers, I'll give it a try when I sort out this
Rockchip boost setup.

On my C630, it never goes higher than 85C with the 4 cores being used, 
and with 2, it never goes about 65C and I do not get any 2.96GHz.  It's 
currently sitting at "6" in the time_in_state for 2965800.


--steev
Thank you for your support.

Regards,
Lukasz

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help