[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6e1fe001-dd21-43fd-ba7b-aae6df819fdd@linuxfoundation.org>
Date: Wed, 18 Jun 2025 16:44:54 -0600
From: Shuah Khan <skhan@...uxfoundation.org>
To: Rohan Lambture <rohanlambture13@...il.com>, trenn@...e.com,
shuah@...nel.org, jwyatt@...hat.com, jkacur@...hat.com
Cc: linux-pm@...r.kernel.org, linux-kernel@...r.kernel.org,
Shuah Khan <skhan@...uxfoundation.org>
Subject: Re: [PATCH] cpupower: Add cpu_start()/cpu_stop() callbacks for
monitors
On 6/9/25 08:49, Rohan Lambture wrote:
> Move per-CPU logic from inside individual monitors to the main
> monitoring framework by adding cpu_start() and cpu_stop() callback
> functions to the cpuidle_monitor structure.
>
> This refactoring allows the framework to handle per-CPU scheduling
> and gives higher priority to fork_it operations as mentioned in
> the TODO. Individual monitors now only need to implement per-CPU
> initialization and cleanup logic without managing the CPU iteration
> themselves.
>
> Changes made:
> - Add cpu_start()/cpu_stop() function pointers to cpuidle_monitor struct
> - Update monitoring framework to call per-CPU callbacks for each CPU
> - Refactor cpuidle_sysfs and mperf monitors to use new callback pattern
> - Maintain backward compatibility for monitors without per-CPU callbacks
>
> This addresses the TODO item: "Add cpu_start()/cpu_stop() callbacks
> for monitor -> This is to move the per_cpu logic from inside the
> monitor to outside it."
Please share details on how this changes has been tested.
Also run checkpatch - there are a few warnings.
>
> Signed-off-by: Rohan Lambture <rohanlambture13@...il.com>
> ---
thanks,
-- Shuah
Powered by blists - more mailing lists