linux-kernel - Re: [PATCH 4/7] cpufreq: intel_pstate: Simplify code in intel_pstate_adjust_busy

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <53973B67.5020802@gmail.com>
Date:	Tue, 10 Jun 2014 10:07:51 -0700
From:	Dirk Brandewie <dirk.brandewie@...il.com>
To:	Stratos Karafotis <stratosk@...aphore.gr>,
	Viresh Kumar <viresh.kumar@...aro.org>
CC:	dirk.brandewie@...il.com, "Rafael J. Wysocki" <rjw@...ysocki.net>,
	Dirk Brandewie <dirk.j.brandewie@...el.com>,
	"linux-pm@...r.kernel.org" <linux-pm@...r.kernel.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 4/7] cpufreq: intel_pstate: Simplify code in intel_pstate_adjust_busy_pstate

On 06/10/2014 07:51 AM, Stratos Karafotis wrote:
> On 10/06/2014 08:27 πμ, Viresh Kumar wrote:
>> On 10 June 2014 02:30, Stratos Karafotis <stratosk@...aphore.gr> wrote:
>>> Simplify the code by removing the inline functions
>>> pstate_increase and pstate_decrease and use directly the
>>> intel_pstate_set_pstate.
>>>

Doesn't apply without your scaled_busy change spin this patch with
out the scaled_busy change and explain the change more fully in the
commit message to cover Viresh's question and I am good with this change.


>>> Signed-off-by: Stratos Karafotis <stratosk@...aphore.gr>
>>> ---
>>>   drivers/cpufreq/intel_pstate.c | 26 +++-----------------------
>>>   1 file changed, 3 insertions(+), 23 deletions(-)
>>>
>>> diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
>>> index 3a49269..26a0262 100644
>>> --- a/drivers/cpufreq/intel_pstate.c
>>> +++ b/drivers/cpufreq/intel_pstate.c
>>> @@ -588,21 +588,6 @@ static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
>>>          pstate_funcs.set(cpu, pstate);
>>>   }
>>>
>>> -static inline void intel_pstate_pstate_increase(struct cpudata *cpu, int steps)
>>> -{
>>> -       int target;
>>> -       target = cpu->pstate.current_pstate + steps;
>>> -
>>> -       intel_pstate_set_pstate(cpu, target);
>>> -}
>>> -
>>> -static inline void intel_pstate_pstate_decrease(struct cpudata *cpu, int steps)
>>> -{
>>> -       int target;
>>> -       target = cpu->pstate.current_pstate - steps;
>>> -       intel_pstate_set_pstate(cpu, target);
>>> -}
>>> -
>>>   static void intel_pstate_get_cpu_pstates(struct cpudata *cpu)
>>>   {
>>>          cpu->pstate.min_pstate = pstate_funcs.get_min();
>>> @@ -695,20 +680,15 @@ static inline void intel_pstate_calc_scaled_busy(struct cpudata *cpu)
>>>   static inline void intel_pstate_adjust_busy_pstate(struct cpudata *cpu)
>>>   {
>>>          struct _pid *pid;
>>> -       signed int ctl = 0;
>>> -       int steps;
>>> +       signed int ctl;
>>>
>>>          pid = &cpu->pid;
>>>          intel_pstate_calc_scaled_busy(cpu);
>>>
>>>          ctl = pid_calc(pid, cpu->sample.busy_scaled);
>>>
>>> -       steps = abs(ctl);
>>> -
>>> -       if (ctl < 0)
>>> -               intel_pstate_pstate_increase(cpu, steps);
>>> -       else
>>> -               intel_pstate_pstate_decrease(cpu, steps);
>>> +       /* Negative values of ctl increase the pstate and vice versa */
>>> +       intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
>>>   }
>>
>> I am not very good at this driver but there is some obvious functional
>> change here. Earlier we used to pass
>> 'cpu->pstate.current_pstate {-|+} steps' and now you are doing '-ctl' only
>>
>
> The original code is:	
>
> 	if (ctl < 0)
> 		intel_pstate_pstate_increase(cpu, steps);
> 	else
> 		intel_pstate_pstate_decrease(cpu, steps);
> 		
> Without inlines functions intel_pstate_pstate_increase() and
> intel_pstate_pstate_decrease() we get:
>
> 	if (ctl < 0)
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate + steps);
> 	else
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - steps);
>
> 	
> But steps = abs(ctl), so:
>
> 	if (ctl < 0)
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate + abs(ctl));
> 	else
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - abs(ctl));
> 	
> By definition, abs(ctl) = ctl if ctl >= 0, -ctl if ctl < 0. Thus:
>
> 	if (ctl < 0)
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate + (-ctl));
> 	else
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
>
> And:
> 	if (ctl < 0)
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
> 	else
> 		intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
>
> Finally remove the unnecessary if statement.		
> 	intel_pstate_set_pstate(cpu, cpu->pstate.current_pstate - ctl);
>
> So, this is equivalent with the original code.
>
> Thanks,
> Stratos
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pm" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/