Re: [Xen-devel] [PATCH 1/2] core-parking: interact with runtime SMT-disabling

[Julien Grall <julien.grall@xxxxxxx>]

Hi Jan,

On 11/04/2019 13:45, Jan Beulich wrote:
> --- a/xen/common/core_parking.c
> +++ b/xen/common/core_parking.c

[...]

> +bool core_parking_remove(unsigned int cpu)
Something looks wrong. This function is implemented in common code but...

> +{
> +    unsigned int i;
> +    bool found = false;
> +
> +    spin_lock(&accounting_lock);
> +
> +    for ( i = 0; i < cur_idle_nums; ++i )
> +        if ( core_parking_cpunum[i] == cpu )
> +        {
> +            found = true;
> +            --cur_idle_nums;
> +            break;
> +        }
> +
> +    for ( ; i < cur_idle_nums; ++i )
> +        core_parking_cpunum[i] = core_parking_cpunum[i + 1];
> +
> +    spin_unlock(&accounting_lock);
> +
> +    return found;
> +}
> +
>   uint32_t get_cur_idle_nums(void)
>   {
>       return cur_idle_nums;
> --- a/xen/include/asm-x86/smp.h
> +++ b/xen/include/asm-x86/smp.h
> @@ -63,6 +63,7 @@ long cpu_up_helper(void *data);
>   long cpu_down_helper(void *data);
>   
>   long core_parking_helper(void *data);
> +bool core_parking_remove(unsigned int cpu);
The prototype is declared in asm-x86/smp.h. Actually, it seems all the function 
exported in core_parking.c have their prototype declared in asm-x86/smp.h.
This raises the question of whether it makes sense to have core_parking.c in 
common. If it makes sense, then the prototype should be declared in include/xen.
Cheers,

--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel