[PATCH] keystone: adds cpu_die implementation

[Vitaly Andrianov]

On 06/29/2015 01:52 PM, Mark Rutland wrote:
> On Mon, Jun 29, 2015 at 06:52:32PM +0100, Vitaly Andrianov wrote:
>> This commit add cpu_die implementation
>>
>> Signed-off-by: Vitaly Andrianov <vitalya at ti.com>
>> ---
>>
>> The discussion of the "keystone: psci: adds cpu_die implementation" commit
>> shows that if PCSI is enabled platform code doesn't need that implementation
>> at all. Having PSCI commands in DTB should be sufficient. Unfortunately
>> Keystone with LPAE enable requires some additional development.
>
> I don't follow.
>
> What do you need to implement for LPAE?
Hi Mark,

The Keystone platform needs to set ttbr1 when it boots secondary core.
It is done in the keystone_smp_secondary_initmem(), which is 
.smp_secondary_init member of the keystone_smp_ops. I couldn't find a 
way how I can add similar function to psci_smp_ops.

Do you have any idea?

>
> Why not implement that rather than adding more platform code that you'll
> likely want to remove later anyway?

If I can solve the above problem, I will not add this code.

Thanks,
Vitaly

>
> Thanks,
> Mark.
>
>> To support HOTPLUG_CPU w/o PSCI we need platform implementation of
>> the cpu_die(), which is added by this patch.
>>
>>   arch/arm/mach-keystone/keystone.h |  1 +
>>   arch/arm/mach-keystone/platsmp.c  | 24 ++++++++++++++++++++++++
>>   2 files changed, 25 insertions(+)
>>
>> diff --git a/arch/arm/mach-keystone/keystone.h b/arch/arm/mach-keystone/keystone.h
>> index cd04a1c..93549cf 100644
>> --- a/arch/arm/mach-keystone/keystone.h
>> +++ b/arch/arm/mach-keystone/keystone.h
>> @@ -12,6 +12,7 @@
>>   #define __KEYSTONE_H__
>>
>>   #define KEYSTONE_MON_CPU_UP_IDX		0x00
>> +#define KEYSTONE_MON_CPU_DIE_IDX	0x84000002
>>
>>   #ifndef __ASSEMBLER__
>>
>> diff --git a/arch/arm/mach-keystone/platsmp.c b/arch/arm/mach-keystone/platsmp.c
>> index 5f46a7c..4e5bdde 100644
>> --- a/arch/arm/mach-keystone/platsmp.c
>> +++ b/arch/arm/mach-keystone/platsmp.c
>> @@ -51,7 +51,31 @@ static inline void __cpuinit keystone_smp_secondary_initmem(unsigned int cpu)
>>   {}
>>   #endif
>>
>> +
>> +#ifdef CONFIG_HOTPLUG_CPU
>> +static void keystone_cpu_die(unsigned int cpu)
>> +{
>> +	int error;
>> +
>> +	pr_debug("keystone-smp: cpu_die %d\n", cpu);
>> +
>> +	error = keystone_cpu_smc(KEYSTONE_MON_CPU_DIE_IDX, cpu, 0);
>> +	if (error)
>> +		pr_err("CPU %d die failed with %d\n", cpu, error);
>> +
>> +	/*
>> +	 * we shouldn't come here. But in case something went
>> +	 * wrong the code below prevents kernel from crush
>> +	 */
>> +	while (1)
>> +		cpu_do_idle();
>> +}
>> +#endif
>> +
>>   struct smp_operations keystone_smp_ops __initdata = {
>>   	.smp_boot_secondary	= keystone_smp_boot_secondary,
>>   	.smp_secondary_init     = keystone_smp_secondary_initmem,
>> +#ifdef CONFIG_HOTPLUG_CPU
>> +	.cpu_die		= keystone_cpu_die,
>> +#endif
>>   };
>> --
>> 1.9.1
>>