Re: [patch 26/34] powerpc: Use generic idle loop

[Deepthi Dharwar <deepthi@linux.vnet.ibm.com>]

On 03/28/2013 09:10 PM, Srivatsa S. Bhat wrote:
> On 03/22/2013 03:23 AM, Thomas Gleixner wrote:
>> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
>> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
>> ---
> [...]
>> Index: linux-2.6/arch/powerpc/kernel/idle.c
>> ===================================================================
>> --- linux-2.6.orig/arch/powerpc/kernel/idle.c
>> +++ linux-2.6/arch/powerpc/kernel/idle.c
>> @@ -50,64 +50,30 @@ static int __init powersave_off(char *ar
>>  }
>>  __setup("powersave=off", powersave_off);
>>
>> -/*
>> - * The body of the idle task.
>> - */
>> -void cpu_idle(void)
>> +void arch_cpu_idle(void)
>>  {
>> -	set_thread_flag(TIF_POLLING_NRFLAG);
>> -	while (1) {
>> -		tick_nohz_idle_enter();
>> -		rcu_idle_enter();
>> -
>> -		while (!need_resched() && !cpu_should_die()) {
>> -			ppc64_runlatch_off();
>> -
>> -			if (ppc_md.power_save) {
>> -				clear_thread_flag(TIF_POLLING_NRFLAG);
>> -				/*
>> -				 * smp_mb is so clearing of TIF_POLLING_NRFLAG
>> -				 * is ordered w.r.t. need_resched() test.
>> -				 */
>> -				smp_mb();
>> -				local_irq_disable();
>> -
>> -				/* Don't trace irqs off for idle */
>> -				stop_critical_timings();
>> -
>> -				/* check again after disabling irqs */
>> -				if (!need_resched() && !cpu_should_die())
>> -					ppc_md.power_save();
>> -
>> -				start_critical_timings();
>> -
>> -				/* Some power_save functions return with
>> -				 * interrupts enabled, some don't.
>> -				 */
>> -				if (irqs_disabled())
>> -					local_irq_enable();
>> -				set_thread_flag(TIF_POLLING_NRFLAG);
>> -
>> -			} else {
>> -				/*
>> -				 * Go into low thread priority and possibly
>> -				 * low power mode.
>> -				 */
>> -				HMT_low();
>> -				HMT_very_low();
>> -			}
>> -		}
>> -
>> -		HMT_medium();
>> -		ppc64_runlatch_on();
>> -		rcu_idle_exit();
>> -		tick_nohz_idle_exit();
>> -		if (cpu_should_die()) {
>> -			sched_preempt_enable_no_resched();
>> -			cpu_die();
>> -		}
>> -		schedule_preempt_disabled();
>> +	ppc64_runlatch_off();
>> +
>> +	if (ppc_md.power_save) {
>> +		ppc_md.power_save();
>> +		/*
>> +		 * Some power_save functions return with
>> +		 * interrupts enabled, some don't.
>> +		 */
>> +		if (irqs_disabled())
>> +			local_irq_enable();
>> +	} else {
>> +		local_irq_enable();
>> +		/*
>> +		 * Go into low thread priority and possibly
>> +		 * low power mode.
>> +		 */
>> +		HMT_low();
>> +		HMT_very_low();
>>  	}
>> +
>> +	HMT_medium();
>> +	ppc64_runlatch_on();
>>  }
>>

The below patch is essential to get hotplug working on POWER.
With that, generic idle loop works fine for PSERIES.

Tested-by: Deepthi Dharwar <deepthi@linux.vnet.ibm.com>

> 
> The call to cpu_die() seems to be missing. And considering this commit,
> 
> commit a7c2bb8279d20d853e43c34584eaf2b039de8026
> Author: Signed-off-by: Darren Hart <dvhltc@us.ibm.com>
> Date:   Wed Aug 4 18:28:33 2010 +0000
> 
>     powerpc: Re-enable preemption before cpu_die()
> 
> we need to call it with preempt enabled. So how about folding the
> following diff with this patch?
> 
> 
> diff --git a/arch/powerpc/kernel/idle.c b/arch/powerpc/kernel/idle.c
> index b751baa..e434fb9 100644
> --- a/arch/powerpc/kernel/idle.c
> +++ b/arch/powerpc/kernel/idle.c
> @@ -33,11 +33,6 @@
>  #include <asm/runlatch.h>
>  #include <asm/smp.h>
> 
> -#ifdef CONFIG_HOTPLUG_CPU
> -#define cpu_should_die()	cpu_is_offline(smp_processor_id())
> -#else
> -#define cpu_should_die()	0
> -#endif
> 
>  unsigned long cpuidle_disable = IDLE_NO_OVERRIDE;
>  EXPORT_SYMBOL(cpuidle_disable);
> @@ -50,6 +45,12 @@ static int __init powersave_off(char *arg)
>  }
>  __setup("powersave=off", powersave_off);
> 
> +void arch_cpu_idle_dead(void)
> +{
> +	sched_preempt_enable_no_resched();
> +	cpu_die();
> +}
> +
>  void arch_cpu_idle(void)
>  {
>  	ppc64_runlatch_off();
> 
> 
> Regards,
> Srivatsa S. Bhat
>