Re: [PATCH 15/21] mips: octeon: smp: Convert to hotplug state machine

[Matt Redfearn <matt.redfearn@imgtec.com>]

HI Sebastian,


On 06/09/16 18:04, Sebastian Andrzej Siewior wrote:
> Install the callbacks via the state machine.
>
> Cc: Ralf Baechle <ralf@linux-mips.org>
> Cc: linux-mips@linux-mips.org
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> ---
>   arch/mips/cavium-octeon/smp.c | 24 +++---------------------
>   include/linux/cpuhotplug.h    |  1 +
>   2 files changed, 4 insertions(+), 21 deletions(-)
>
> diff --git a/arch/mips/cavium-octeon/smp.c b/arch/mips/cavium-octeon/smp.c
> index 4d457d602d3b..228c1cab2912 100644
> --- a/arch/mips/cavium-octeon/smp.c
> +++ b/arch/mips/cavium-octeon/smp.c
> @@ -380,29 +380,11 @@ static int octeon_update_boot_vector(unsigned int cpu)
>   	return 0;
>   }
>   
> -static int octeon_cpu_callback(struct notifier_block *nfb,
> -	unsigned long action, void *hcpu)
> -{
> -	unsigned int cpu = (unsigned long)hcpu;
> -
> -	switch (action & ~CPU_TASKS_FROZEN) {
> -	case CPU_UP_PREPARE:
> -		octeon_update_boot_vector(cpu);
> -		break;
> -	case CPU_ONLINE:
> -		pr_info("Cpu %d online\n", cpu);
> -		break;
> -	case CPU_DEAD:
> -		break;
> -	}
> -
> -	return NOTIFY_OK;
> -}
> -
>   static int register_cavium_notifier(void)
>   {
> -	hotcpu_notifier(octeon_cpu_callback, 0);
> -	return 0;
> +	return cpuhp_setup_state_nocalls(CPUHP_MIPS_CAVIUM_PREPARE,
> +					 "mips/cavium:prepare",
> +					 octeon_update_boot_vector, NULL);

This looks like a nice change which results in much cleaner code

>   }
>   late_initcall(register_cavium_notifier);
>   
> diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h
> index 4cbf88c955d6..058d312fdf6f 100644
> --- a/include/linux/cpuhotplug.h
> +++ b/include/linux/cpuhotplug.h
> @@ -44,6 +44,7 @@ enum cpuhp_state {
>   	CPUHP_SH_SH3X_PREPARE,
>   	CPUHP_X86_MICRCODE_PREPARE,
>   	CPUHP_NOTF_ERR_INJ_PREPARE,
> +	CPUHP_MIPS_CAVIUM_PREPARE,

But I'm curious why we have to create a new state here - this is going 
to get very unwieldy if every variant of every architecture has to have 
it's own state values in that enumeration. Can this use, what I assume 
is (and perhaps could be documented better in 
include/linux/cpuhotplug.h), the generic prepare state CPUHP_NOTIFY_PREPARE?

Thanks,
Matt

>   	CPUHP_TIMERS_DEAD,
>   	CPUHP_BRINGUP_CPU,
>   	CPUHP_AP_IDLE_DEAD,