Re: [PATCH] x86: Don't copy per_cpu cpuinfo for BSP two times

[Borislav Petkov <bp@alien8.de>]

On Fri, Jan 21, 2011 at 03:29:54PM -0800, Yinghai Lu wrote:
> 
> smp_store_cpu_info(0) will do that.
> 
> Signed-off-by: Yinghai Lu <yinghai@kernel.org>
> 
> ---
>  arch/x86/kernel/smpboot.c |    6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> Index: linux-2.6/arch/x86/kernel/smpboot.c
> ===================================================================
> --- linux-2.6.orig/arch/x86/kernel/smpboot.c
> +++ linux-2.6/arch/x86/kernel/smpboot.c
> @@ -1071,13 +1071,13 @@ void __init native_smp_prepare_cpus(unsi
>  
>  	preempt_disable();
>  	smp_cpu_index_default();
> -	memcpy(__this_cpu_ptr(&cpu_info), &boot_cpu_data, sizeof(cpu_info));
> -	cpumask_copy(cpu_callin_mask, cpumask_of(0));
> -	mb();
> +
>  	/*
>  	 * Setup boot CPU information
>  	 */
>  	smp_store_cpu_info(0); /* Final full version of the data */
> +	cpumask_copy(cpu_callin_mask, cpumask_of(0));
> +	mb();

Question: why do we need the memory barrier there? We're running on the
BSP and long before smp_init() to be all careful about the callin_mask.
It could be there as a left over from the 32 and 64-bit merge, AFAICT
from git history...

Thanks.

-- 
Regards/Gruss,
    Boris.