All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ionela Voinescu <ionela.voinescu@arm.com>
To: Lukasz Luba <lukasz.luba@arm.com>
Cc: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
	dri-devel@lists.freedesktop.org, amit.kucheria@verdurent.com,
	airlied@linux.ie, daniel.lezcano@linaro.org,
	steven.price@arm.com, alyssa.rosenzweig@collabora.com,
	rui.zhang@intel.com, orjan.eide@arm.com
Subject: Re: [PATCH 2/5] thermal: devfreq_cooling: get a copy of device status
Date: Wed, 7 Oct 2020 17:11:20 +0100	[thread overview]
Message-ID: <20201007161120.GC15063@arm.com> (raw)
In-Reply-To: <20200921122007.29610-3-lukasz.luba@arm.com>

On Monday 21 Sep 2020 at 13:20:04 (+0100), Lukasz Luba wrote:
> Devfreq cooling needs to now the correct status of the device in order
> to operate. Do not rely on Devfreq last_status which might be a stale data
> and get more up-to-date values of the load.
> 
> Devfreq framework can change the device status in the background. To
> mitigate this situation make a copy of the status structure and use it
> for internal calculations.
> 
> In addition this patch adds normalization function, which also makes sure
> that whatever data comes from the device, it is in a sane range.
> 
> Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
> ---
>  drivers/thermal/devfreq_cooling.c | 52 +++++++++++++++++++++++++------
>  1 file changed, 43 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/thermal/devfreq_cooling.c b/drivers/thermal/devfreq_cooling.c
> index 7063ccb7b86d..cf045bd4d16b 100644
> --- a/drivers/thermal/devfreq_cooling.c
> +++ b/drivers/thermal/devfreq_cooling.c
> @@ -227,6 +227,24 @@ static inline unsigned long get_total_power(struct devfreq_cooling_device *dfc,
>  							       voltage);
>  }
>  
> +static void _normalize_load(struct devfreq_dev_status *status)

Is there a reason for the leading "_" ?
AFAIK, "__name()" is meant to suggest a "worker" function for another
"name()" function, but that would not apply here.

> +{
> +	/* Make some space if needed */
> +	if (status->busy_time > 0xffff) {
> +		status->busy_time >>= 10;
> +		status->total_time >>= 10;
> +	}

How about removing the above code and adding here:

status->busy_time = status->busy_time ? : 1;

> +
> +	if (status->busy_time > status->total_time)

This check would then cover the possibility that total_time is 0.

> +		status->busy_time = status->total_time;

But a reversal is needed here:
		status->total_time = status->busy_time;

> +
> +	status->busy_time *= 100;
> +	status->busy_time /= status->total_time ? : 1;
> +
> +	/* Avoid division by 0 */
> +	status->busy_time = status->busy_time ? : 1;
> +	status->total_time = 100;

Then all of this code can be replaced by:

status->busy_time = (unsigned long)div64_u64((u64)status->busy_time << 10,
					     status->total_time);
status->total_time = 1 << 10;

This way you gain some resolution to busy_time and the divisions in the
callers would just become shifts by 10.

Hope it helps,
Ionela.

WARNING: multiple messages have this Message-ID (diff)
From: Ionela Voinescu <ionela.voinescu@arm.com>
To: Lukasz Luba <lukasz.luba@arm.com>
Cc: amit.kucheria@verdurent.com, linux-pm@vger.kernel.org,
	airlied@linux.ie, daniel.lezcano@linaro.org,
	linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org,
	steven.price@arm.com, alyssa.rosenzweig@collabora.com,
	rui.zhang@intel.com, orjan.eide@arm.com
Subject: Re: [PATCH 2/5] thermal: devfreq_cooling: get a copy of device status
Date: Wed, 7 Oct 2020 17:11:20 +0100	[thread overview]
Message-ID: <20201007161120.GC15063@arm.com> (raw)
In-Reply-To: <20200921122007.29610-3-lukasz.luba@arm.com>

On Monday 21 Sep 2020 at 13:20:04 (+0100), Lukasz Luba wrote:
> Devfreq cooling needs to now the correct status of the device in order
> to operate. Do not rely on Devfreq last_status which might be a stale data
> and get more up-to-date values of the load.
> 
> Devfreq framework can change the device status in the background. To
> mitigate this situation make a copy of the status structure and use it
> for internal calculations.
> 
> In addition this patch adds normalization function, which also makes sure
> that whatever data comes from the device, it is in a sane range.
> 
> Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
> ---
>  drivers/thermal/devfreq_cooling.c | 52 +++++++++++++++++++++++++------
>  1 file changed, 43 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/thermal/devfreq_cooling.c b/drivers/thermal/devfreq_cooling.c
> index 7063ccb7b86d..cf045bd4d16b 100644
> --- a/drivers/thermal/devfreq_cooling.c
> +++ b/drivers/thermal/devfreq_cooling.c
> @@ -227,6 +227,24 @@ static inline unsigned long get_total_power(struct devfreq_cooling_device *dfc,
>  							       voltage);
>  }
>  
> +static void _normalize_load(struct devfreq_dev_status *status)

Is there a reason for the leading "_" ?
AFAIK, "__name()" is meant to suggest a "worker" function for another
"name()" function, but that would not apply here.

> +{
> +	/* Make some space if needed */
> +	if (status->busy_time > 0xffff) {
> +		status->busy_time >>= 10;
> +		status->total_time >>= 10;
> +	}

How about removing the above code and adding here:

status->busy_time = status->busy_time ? : 1;

> +
> +	if (status->busy_time > status->total_time)

This check would then cover the possibility that total_time is 0.

> +		status->busy_time = status->total_time;

But a reversal is needed here:
		status->total_time = status->busy_time;

> +
> +	status->busy_time *= 100;
> +	status->busy_time /= status->total_time ? : 1;
> +
> +	/* Avoid division by 0 */
> +	status->busy_time = status->busy_time ? : 1;
> +	status->total_time = 100;

Then all of this code can be replaced by:

status->busy_time = (unsigned long)div64_u64((u64)status->busy_time << 10,
					     status->total_time);
status->total_time = 1 << 10;

This way you gain some resolution to busy_time and the divisions in the
callers would just become shifts by 10.

Hope it helps,
Ionela.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

  reply	other threads:[~2020-10-07 16:11 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-21 12:20 [PATCH 0/5] Thermal devfreq cooling improvements with Energy Model Lukasz Luba
2020-09-21 12:20 ` Lukasz Luba
2020-09-21 12:20 ` [PATCH 1/5] thermal: devfreq_cooling: change tracing function and arguments Lukasz Luba
2020-09-21 12:20   ` Lukasz Luba
2020-09-21 12:20 ` [PATCH 2/5] thermal: devfreq_cooling: get a copy of device status Lukasz Luba
2020-09-21 12:20   ` Lukasz Luba
2020-10-07 16:11   ` Ionela Voinescu [this message]
2020-10-07 16:11     ` Ionela Voinescu
2020-10-22 10:55     ` Lukasz Luba
2020-10-22 10:55       ` Lukasz Luba
2020-12-01 10:36       ` Ionela Voinescu
2020-12-01 10:36         ` Ionela Voinescu
2020-12-01 12:19         ` Lukasz Luba
2020-12-01 12:19           ` Lukasz Luba
2020-12-01 14:55           ` Ionela Voinescu
2020-12-01 14:55             ` Ionela Voinescu
2020-10-14 14:34   ` Daniel Lezcano
2020-10-14 14:34     ` Daniel Lezcano
2020-10-22 11:45     ` Lukasz Luba
2020-10-22 11:45       ` Lukasz Luba
2020-09-21 12:20 ` [PATCH 3/5] thermal: devfreq_cooling: add new registration functions with Energy Model Lukasz Luba
2020-09-21 12:20   ` Lukasz Luba
2020-10-07 12:07   ` Ionela Voinescu
2020-10-07 12:07     ` Ionela Voinescu
2020-10-22 11:17     ` Lukasz Luba
2020-10-22 11:17       ` Lukasz Luba
2020-12-01 14:05       ` Ionela Voinescu
2020-12-01 14:05         ` Ionela Voinescu
2020-12-01 14:37         ` Lukasz Luba
2020-12-01 14:37           ` Lukasz Luba
2020-12-01 15:02           ` Ionela Voinescu
2020-12-01 15:02             ` Ionela Voinescu
2020-09-21 12:20 ` [PATCH 4/5] thermal: devfreq_cooling: remove old power model and use EM Lukasz Luba
2020-09-21 12:20   ` Lukasz Luba
2020-10-07 15:12   ` Ionela Voinescu
2020-10-07 15:12     ` Ionela Voinescu
2020-10-22 11:26     ` Lukasz Luba
2020-10-22 11:26       ` Lukasz Luba
2020-09-21 12:20 ` [PATCH 5/5] drm/panfrost: Register devfreq cooling and attempt to add Energy Model Lukasz Luba
2020-09-21 12:20   ` 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=20201007161120.GC15063@arm.com \
    --to=ionela.voinescu@arm.com \
    --cc=airlied@linux.ie \
    --cc=alyssa.rosenzweig@collabora.com \
    --cc=amit.kucheria@verdurent.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=lukasz.luba@arm.com \
    --cc=orjan.eide@arm.com \
    --cc=rui.zhang@intel.com \
    --cc=steven.price@arm.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.