Re: [PATCH 7/7] perf/x86/intel/P4: Fix smp_num_siblings usage

From: Peter Zijlstra
Date: Mon Aug 15 2022 - 05:12:35 EST


On Sat, Aug 13, 2022 at 12:41:44AM +0800, Zhang Rui wrote:
> smp_num_siblings can be larger than 2.

Not on a P4 it can't ;-)

>
> Any value larger than 1 suggests HT is supported.
>
> Reviewed-by: Len Brown <len.brown@xxxxxxxxx>
> Signed-off-by: Zhang Rui <rui.zhang@xxxxxxxxx>
> ---
> arch/x86/include/asm/perf_event_p4.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h
> index 94de1a05aeba..b14e9a20a7c0 100644
> --- a/arch/x86/include/asm/perf_event_p4.h
> +++ b/arch/x86/include/asm/perf_event_p4.h
> @@ -189,7 +189,7 @@ static inline int p4_ht_active(void)
> static inline int p4_ht_thread(int cpu)
> {
> #ifdef CONFIG_SMP
> - if (smp_num_siblings == 2)
> + if (smp_num_siblings > 1)
> return cpu != cpumask_first(this_cpu_cpumask_var_ptr(cpu_sibling_map));
> #endif
> return 0;

Unless Intel plans to respin an P4 with extra siblings on, I don't think
this qualifies for the word 'fix'.