Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1499213 > unrolled thread

[PATCH] cpufreq: CPPC: Correct desired_perf calculation

Started byHoan Tran <hotran@apm.com>
First post2016-10-11 23:40 +0200
Last post2016-10-12 19:40 +0200
Articles 6 — 4 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] cpufreq: CPPC: Correct desired_perf calculation Hoan Tran <hotran@apm.com> - 2016-10-11 23:40 +0200
    Re: [PATCH] cpufreq: CPPC: Correct desired_perf calculation "Rafael J. Wysocki" <rjw@rjwysocki.net> - 2016-10-11 23:50 +0200
      Re: [PATCH] cpufreq: CPPC: Correct desired_perf calculation Hoan Tran <hotran@apm.com> - 2016-10-12 00:40 +0200
    Re: [PATCH] cpufreq: CPPC: Correct desired_perf calculation Viresh Kumar <viresh.kumar@linaro.org> - 2016-10-12 05:30 +0200
    Re: [PATCH] cpufreq: CPPC: Correct desired_perf calculation "Prakash, Prashanth" <pprakash@codeaurora.org> - 2016-10-12 17:40 +0200
      Re: [PATCH] cpufreq: CPPC: Correct desired_perf calculation Hoan Tran <hotran@apm.com> - 2016-10-12 19:40 +0200

#1499213 — [PATCH] cpufreq: CPPC: Correct desired_perf calculation

FromHoan Tran <hotran@apm.com>
Date2016-10-11 23:40 +0200
Subject[PATCH] cpufreq: CPPC: Correct desired_perf calculation
Message-ID<srcWd-3tt-9@gated-at.bofh.it>
The desired_perf is an abstract performance number. Its value should
be in the range of [lowest perf, highest perf] of CPPC.
The correct calculation is
  desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz

Signed-off-by: Hoan Tran <hotran@apm.com>
---
 drivers/cpufreq/cppc_cpufreq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
index 1b2f28f..ab1d4b7 100644
--- a/drivers/cpufreq/cppc_cpufreq.c
+++ b/drivers/cpufreq/cppc_cpufreq.c
@@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
 
 	cpu = all_cpu_data[policy->cpu];
 
-	cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
+	cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
 	freqs.old = policy->cur;
 	freqs.new = target_freq;
 
-- 
1.9.1

[toc] | [next] | [standalone]

#1499215

From"Rafael J. Wysocki" <rjw@rjwysocki.net>
Date2016-10-11 23:50 +0200
Message-ID<srd5T-3x4-15@gated-at.bofh.it>
In reply to#1499213
On Tuesday, October 11, 2016 02:12:00 PM Hoan Tran wrote:
> The desired_perf is an abstract performance number. Its value should
> be in the range of [lowest perf, highest perf] of CPPC.
> The correct calculation is
>   desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz
> 
> Signed-off-by: Hoan Tran <hotran@apm.com>
> ---
>  drivers/cpufreq/cppc_cpufreq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
> index 1b2f28f..ab1d4b7 100644
> --- a/drivers/cpufreq/cppc_cpufreq.c
> +++ b/drivers/cpufreq/cppc_cpufreq.c
> @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
>  
>  	cpu = all_cpu_data[policy->cpu];
>  
> -	cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
> +	cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
>  	freqs.old = policy->cur;
>  	freqs.new = target_freq;
>  
> 

This patch has already been applied AFAICS.

Thanks,
Rafael

[toc] | [prev] | [next] | [standalone]

#1499278

FromHoan Tran <hotran@apm.com>
Date2016-10-12 00:40 +0200
Message-ID<srdSh-47Y-1@gated-at.bofh.it>
In reply to#1499215
Hi Rafael,

On Tue, Oct 11, 2016 at 2:50 PM, Rafael J. Wysocki <rjw@rjwysocki.net> wrote:
> On Tuesday, October 11, 2016 02:12:00 PM Hoan Tran wrote:
>> The desired_perf is an abstract performance number. Its value should
>> be in the range of [lowest perf, highest perf] of CPPC.
>> The correct calculation is
>>   desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz
>>
>> Signed-off-by: Hoan Tran <hotran@apm.com>
>> ---
>>  drivers/cpufreq/cppc_cpufreq.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
>> index 1b2f28f..ab1d4b7 100644
>> --- a/drivers/cpufreq/cppc_cpufreq.c
>> +++ b/drivers/cpufreq/cppc_cpufreq.c
>> @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
>>
>>       cpu = all_cpu_data[policy->cpu];
>>
>> -     cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
>> +     cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
>>       freqs.old = policy->cur;
>>       freqs.new = target_freq;
>>
>>
>
> This patch has already been applied AFAICS.
>

You mean this applied patch "cpufreq: CPPC: Avoid overflow when
calculating desired_perf"

This is another the bug, not the overflow bug.

Thanks
Hoan



> Thanks,
> Rafael
>

[toc] | [prev] | [next] | [standalone]

#1499336

FromViresh Kumar <viresh.kumar@linaro.org>
Date2016-10-12 05:30 +0200
Message-ID<srioV-78a-1@gated-at.bofh.it>
In reply to#1499213
On 11-10-16, 14:12, Hoan Tran wrote:
> The desired_perf is an abstract performance number. Its value should
> be in the range of [lowest perf, highest perf] of CPPC.
> The correct calculation is
>   desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz
> 
> Signed-off-by: Hoan Tran <hotran@apm.com>
> ---
>  drivers/cpufreq/cppc_cpufreq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
> index 1b2f28f..ab1d4b7 100644
> --- a/drivers/cpufreq/cppc_cpufreq.c
> +++ b/drivers/cpufreq/cppc_cpufreq.c
> @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
>  
>  	cpu = all_cpu_data[policy->cpu];
>  
> -	cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
> +	cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
>  	freqs.old = policy->cur;
>  	freqs.new = target_freq;

I am not sure what the calculations should be like, but I don't have
any objections against applying this..

Acked-by: Viresh Kumar <viresh.kumar@linaro.org>

-- 
viresh

[toc] | [prev] | [next] | [standalone]

#1499778

From"Prakash, Prashanth" <pprakash@codeaurora.org>
Date2016-10-12 17:40 +0200
Message-ID<srtNo-6jC-19@gated-at.bofh.it>
In reply to#1499213
Hi Hoan,

On 10/11/2016 3:12 PM, Hoan Tran wrote:
> The desired_perf is an abstract performance number. Its value should
> be in the range of [lowest perf, highest perf] of CPPC.
> The correct calculation is
>   desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz
>
> Signed-off-by: Hoan Tran <hotran@apm.com>
> ---
>  drivers/cpufreq/cppc_cpufreq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
> index 1b2f28f..ab1d4b7 100644
> --- a/drivers/cpufreq/cppc_cpufreq.c
> +++ b/drivers/cpufreq/cppc_cpufreq.c
> @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
>  
>  	cpu = all_cpu_data[policy->cpu];
>  
> -	cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
> +	cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
The patch looks good, I suppose we can add a small optimization. We can do a simple check
to see if the newly computed desired_perf is same as old one, If it is same we can just return
here instead of calling cppc_set_perf with same desired_perf value.

Thanks,
Prashanth

[toc] | [prev] | [next] | [standalone]

#1499846

FromHoan Tran <hotran@apm.com>
Date2016-10-12 19:40 +0200
Message-ID<srvFv-7we-7@gated-at.bofh.it>
In reply to#1499778
[Resend with plain text mode]

Hi Prashanth,

On Wed, Oct 12, 2016 at 8:36 AM, Prakash, Prashanth
<pprakash@codeaurora.org> wrote:
> Hi Hoan,
>
> On 10/11/2016 3:12 PM, Hoan Tran wrote:
>> The desired_perf is an abstract performance number. Its value should
>> be in the range of [lowest perf, highest perf] of CPPC.
>> The correct calculation is
>>   desired_perf = freq * cppc_highest_perf / cppc_dmi_max_khz
>>
>> Signed-off-by: Hoan Tran <hotran@apm.com>
>> ---
>>  drivers/cpufreq/cppc_cpufreq.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
>> index 1b2f28f..ab1d4b7 100644
>> --- a/drivers/cpufreq/cppc_cpufreq.c
>> +++ b/drivers/cpufreq/cppc_cpufreq.c
>> @@ -84,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy,
>>
>>       cpu = all_cpu_data[policy->cpu];
>>
>> -     cpu->perf_ctrls.desired_perf = (u64)target_freq * policy->max / cppc_dmi_max_khz;
>> +     cpu->perf_ctrls.desired_perf = (u64)target_freq * cpu->perf_caps.highest_perf / cppc_dmi_max_khz;
> The patch looks good, I suppose we can add a small optimization. We can do a simple check
> to see if the newly computed desired_perf is same as old one, If it is same we can just return
> here instead of calling cppc_set_perf with same desired_perf value.

That's a good point. I can add a check into this patch.

Thanks
Hoan

>
> Thanks,
> Prashanth

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web