Re: [PATCH v4 05/24] x86: refactor psr: implement Domain init/free and schedule flows.

["Jan Beulich" <JBeulich@suse.com>]

>>> On 14.12.16 at 05:07, <yi.y.sun@linux.intel.com> wrote:
> @@ -358,11 +366,32 @@ void psr_free_rmid(struct domain *d)
>      d->arch.psr_rmid = 0;
>  }
>  
> +static inline unsigned int get_max_cos_max(const struct psr_socket_info *info)
> +{
> +    const struct feat_node *feat_tmp;

Same remark as on the earlier patch regarding the _tmp suffix.
Should this reoccur in later patches, I won't give the same
comment again.

> +    unsigned int cos_max = 0;
> +
> +    list_for_each_entry(feat_tmp, &info->feat_list, list)
> +        cos_max = max(feat_tmp->ops.get_max_cos_max(feat_tmp), cos_max);
> +
> +    return cos_max;
> +}
> +
>  static inline void psr_assoc_init(void)
>  {
>      struct psr_assoc *psra = &this_cpu(psr_assoc);
>  
> -    if ( psr_cmt_enabled() )
> +    if ( socket_info )
> +    {
> +        unsigned int socket = cpu_to_socket(smp_processor_id());
> +        const struct psr_socket_info *info = socket_info + socket;
> +        unsigned int cos_max = get_max_cos_max(info);
> +
> +        if ( info->feat_mask )
> +            psra->cos_mask = ((1ull << get_count_order(cos_max)) - 1) << 32;
> +    }
> +
> +    if ( psr_cmt_enabled() || psra->cos_mask )
>          rdmsrl(MSR_IA32_PSR_ASSOC, psra->val);
>  }
>  
> @@ -371,6 +400,12 @@ static inline void psr_assoc_rmid(uint64_t *reg, unsigned int rmid)
>      *reg = (*reg & ~rmid_mask) | (rmid & rmid_mask);
>  }
>  
> +static inline void psr_assoc_cos(uint64_t *reg, unsigned int cos,
> +                                 uint64_t cos_mask)
> +{
> +    *reg = (*reg & ~cos_mask) | (((uint64_t)cos << 32) & cos_mask);

This recurring 32 would perhaps better become a #define, so they
can be identified as the same entity (at least I think they are).

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel