Re: [RFC/PATCH 4/6] ARM: shmobile: r8a7791: Add Core-Standby CPUIdle support for r8a7791

[keita kobayashi <keita.kobayashi.ym@renesas.com>]

Hi Manish, Simon

Thank you for your comment.

(2015/03/05 9:29), Simon Horman wrote:

> On Wed, Mar 04, 2015 at 11:03:03AM +0530, Manish Badarkhe wrote:
>> Hi,
>>
>> On Wed, Mar 4, 2015 at 10:25 AM, Keita Kobayashi <
>> keita.kobayashi.ym@renesas.com> wrote:
>>
>>> This patch Add the wfi and Core-Standby for CPUIdle power states
>>> support for the r8a7791 SoC.
>>>
>>> Signed-off-by: Keita Kobayashi <keita.kobayashi.ym@renesas.com>
>>> ---
>>>  arch/arm/mach-shmobile/pm-r8a7791.c | 41
>>> +++++++++++++++++++++++++++++++++++++
>>>  1 file changed, 41 insertions(+)
>>>
>>> diff --git a/arch/arm/mach-shmobile/pm-r8a7791.c
>>> b/arch/arm/mach-shmobile/pm-r8a7791.c
>>> index 25f107b..6c00d37 100644
>>> --- a/arch/arm/mach-shmobile/pm-r8a7791.c
>>> +++ b/arch/arm/mach-shmobile/pm-r8a7791.c
>>> @@ -10,8 +10,13 @@
>>>   * for more details.
>>>   */
>>>
>>> +#include <linux/cpuidle.h>
>>>  #include <linux/kernel.h>
>>>  #include <linux/smp.h>
>>> +#include <linux/platform_data/renesas-cpuidle.h>
>>> +#include <linux/pm.h>
>>> +#include <asm/cpuidle.h>
>>> +#include <asm/cputype.h>
>>>  #include <asm/io.h>
>>>  #include "common.h"
>>>  #include "pm-rcar.h"
>>> @@ -43,6 +48,41 @@ static inline void r8a7791_sysc_init(void) {}
>>>
>>>  #endif /* CONFIG_SMP */
>>>
>>> +#ifdef CONFIG_ARM_RENESAS_CPUIDLE
>>> +static struct cpuidle_driver r8a7791_cpuidle_driver = {
>>> +       .name = "rcar_r8a7791_cpuidle",
>>> +       .owner = THIS_MODULE,
>>> +       .states = {
>>> +               [0] = ARM_CPUIDLE_WFI_STATE,
>>> +               [1] = {
>>> +                       .name = "Core-Standby",
>>> +                       .desc = "Core Standby Mode",
>>> +                       .exit_latency = 2800,
>>> +                       .target_residency = 2800,
>>> +                       .enter = shmobile_smp_apmu_enter_cpuidle,
>>> +               },
>>> +       },
>>> +       .state_count = 2,
>>> +};
>>> +
>>> +static struct renesas_cpuidle_driver r8a7791_cpuidle = {
>>> +       .target_cpu     = ARM_CPU_PART_CORTEX_A15,
>>> +       .renesas_drv    = &r8a7791_cpuidle_driver,
>>> +};
>>> +
>>> +static struct renesas_cpuidle r8a7791_data = {
>>> +       .num_drvs       = 1,
>>> +       .rcd            = &r8a7791_cpuidle,
>>> +};
>>> +
>>> +static void __init r8a7791_cpuidle_init(void)
>>> +{
>>> +       shmobile_cpuidle_set_driver(&r8a7791_data);
>>> +}
>>> +#else
>>> +static void __init r8a7791_cpuidle_init(void) {}
>>> +#endif
>>>
>>
>> You can use 'IS_ENABLED' instead of this '#ifdef' during calling of
>> function 'r8a7791_cpuidle_init' to avoid empty function in '#else' part
>> of the code.
>>
>> May be it is possible to replace all '#ifdef' with 'IS_ENABLED' call
>> in this file and also rest of patches in this series.
> 
> Kobayashi-san,
> 
> it seems likely that Manish's suggestion will likely lead
> to cleaner code. Could you look into it?

I will fix it with V2 patch series.

Regards.
Keita Kobayashi