linux-pm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dietmar Eggemann <dietmar.eggemann@arm.com>
To: Lukasz Luba <lukasz.luba@arm.com>,
	linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
	rafael@kernel.org
Cc: rui.zhang@intel.com, amit.kucheria@verdurent.com,
	amit.kachhap@gmail.com, daniel.lezcano@linaro.org,
	viresh.kumar@linaro.org, len.brown@intel.com, pavel@ucw.cz,
	Pierre.Gondois@arm.com, ionela.voinescu@arm.com,
	rostedt@goodmis.org, mhiramat@kernel.org
Subject: Re: [PATCH v2 08/17] PM: EM: Introduce runtime modifiable table
Date: Tue, 30 May 2023 12:18:21 +0200	[thread overview]
Message-ID: <c4052a3d-ab24-6c20-ed07-d752098e6286@arm.com> (raw)
In-Reply-To: <20230512095743.3393563-9-lukasz.luba@arm.com>

On 12/05/2023 11:57, Lukasz Luba wrote:
> This patch introduces the new feature: modifiable EM perf_state table.
> The new runtime table would be populated with a new power data to better
> reflect the actual power. The power can vary over time e.g. due to the
> SoC temperature change. Higher temperature can increase power values.
> For longer running scenarios, such as game or camera, when also other
> devices are used (e.g. GPU, ISP) the CPU power can change. The new
> EM framework is able to addresses this issue and change the data
> at runtime safely. The runtime modifiable EM data is used by the Energy
> Aware Scheduler (EAS) for the task placement.

It's important to say that EAS is the _only_user of the `runtime
modifiable EM`. All the other users (thermal, etc.) are still using the
default (basic) EM. IMHO, this fact drove the design here.

> Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
> ---
>  include/linux/energy_model.h | 13 +++++++++++++
>  kernel/power/energy_model.c  | 24 ++++++++++++++++++++++++
>  2 files changed, 37 insertions(+)
> 
> diff --git a/include/linux/energy_model.h b/include/linux/energy_model.h
> index cc2bf607191e..a616006a8130 100644
> --- a/include/linux/energy_model.h
> +++ b/include/linux/energy_model.h
> @@ -36,9 +36,21 @@ struct em_perf_state {
>   */
>  #define EM_PERF_STATE_INEFFICIENT BIT(0)
>  
> +/**
> + * struct em_perf_table - Performance states table, which can be
> + *		runtime modifiable and protected with RCU

which is `runtime modifiable` ? So `runtime modifiable performance state
table`? RCU is obvious since we have `struct rcu_head rcu`.

> + * @state:	List of performance states, in ascending order
> + * @rcu:	RCU used for safe access and destruction
> + */
> +struct em_perf_table {
> +	struct em_perf_state *state;
> +	struct rcu_head rcu;
> +};
> +
>  /**
>   * struct em_perf_domain - Performance domain
>   * @table:		List of performance states, in ascending order
> + * @runtime_table:	Pointer to the runtime modified em_perf_table

s/modified/modifiable

[...]

> @@ -237,12 +238,23 @@ static int em_create_pd(struct device *dev, int nr_states,
>  			return -ENOMEM;
>  	}
>  
> +	runtime_table = kzalloc(sizeof(*runtime_table), GFP_KERNEL);
> +	if (!runtime_table) {
> +		kfree(pd);
> +		return -ENOMEM;
> +	}
> +
>  	ret = em_create_perf_table(dev, pd, nr_states, cb, flags);
>  	if (ret) {
>  		kfree(pd);
> +		kfree(runtime_table);
>  		return ret;
>  	}
>  
> +	/* Re-use temporally (till 1st modification) the memory */

So this means that the runtime (modifiable) table
(pd->runtime_table>state) is mapped to the default (basic) table
(pd->default_table->state) until the first call to
em_dev_update_perf_domain() (here mentioned as the 1st modification)?

IMHO, not easy to understand since neither the cover letter, nor
documentation patch 15/17 describes this in a consistent story.

[...]

  parent reply	other threads:[~2023-05-30 10:18 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-12  9:57 [PATCH v2 00/17] Introduce runtime modifiable Energy Model Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 01/17] PM: EM: Refactor em_cpufreq_update_efficiencies() arguments Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 02/17] PM: EM: Find first CPU online while updating OPP efficiency Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 03/17] PM: EM: Refactor em_pd_get_efficient_state() to be more flexible Lukasz Luba
2023-05-30 11:06   ` Dietmar Eggemann
2023-07-03 16:22     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 04/17] PM: EM: Create a new function em_compute_costs() Lukasz Luba
2023-05-30  9:51   ` Dietmar Eggemann
2023-07-03 15:09     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 05/17] trace: energy_model: Add trace event for EM runtime modifications Lukasz Luba
2023-05-30 10:03   ` Dietmar Eggemann
2023-07-03 15:53     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 06/17] PM: EM: Add update_power() callback for " Lukasz Luba
2023-05-30  9:31   ` Dietmar Eggemann
2023-07-03 15:06     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 07/17] PM: EM: Check if the get_cost() callback is present in em_compute_costs() Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 08/17] PM: EM: Introduce runtime modifiable table Lukasz Luba
2023-05-14  4:28   ` kernel test robot
2023-05-30 10:18   ` Dietmar Eggemann [this message]
2023-07-03 15:58     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 09/17] PM: EM: Add RCU mechanism which safely cleans the old data Lukasz Luba
2023-05-30 10:02   ` Dietmar Eggemann
2023-07-03 15:49     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 10/17] PM: EM: Add runtime update interface to modify EM power Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 11/17] PM: EM: Use runtime modified EM for CPUs energy estimation in EAS Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 12/17] PM: EM: Add argument to get_cost() for runtime modification Lukasz Luba
2023-05-30  9:53   ` Dietmar Eggemann
2023-07-03 15:30     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 13/17] PM: EM: Refactor struct em_perf_domain and add default_table Lukasz Luba
2023-05-30 10:23   ` Dietmar Eggemann
2023-07-03 16:00     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 14/17] Documentation: EM: Add a new section about the design Lukasz Luba
2023-05-30 10:33   ` Dietmar Eggemann
2023-07-03 16:09     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 15/17] Documentation: EM: Add a runtime modifiable EM design description Lukasz Luba
2023-05-30 10:42   ` Dietmar Eggemann
2023-07-03 16:13     ` Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 16/17] Documentation: EM: Add example with driver modifying the EM Lukasz Luba
2023-05-12  9:57 ` [PATCH v2 17/17] Documentation: EM: Describe the API of runtime modifications Lukasz Luba
2023-05-24 17:25 ` [PATCH v2 00/17] Introduce runtime modifiable Energy Model Rafael J. Wysocki
2023-07-03 11:08   ` Lukasz Luba
2023-05-30 11:07 ` Dietmar Eggemann
2023-07-03 16:35   ` Lukasz Luba

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c4052a3d-ab24-6c20-ed07-d752098e6286@arm.com \
    --to=dietmar.eggemann@arm.com \
    --cc=Pierre.Gondois@arm.com \
    --cc=amit.kachhap@gmail.com \
    --cc=amit.kucheria@verdurent.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=ionela.voinescu@arm.com \
    --cc=len.brown@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=lukasz.luba@arm.com \
    --cc=mhiramat@kernel.org \
    --cc=pavel@ucw.cz \
    --cc=rafael@kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=rui.zhang@intel.com \
    --cc=viresh.kumar@linaro.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).