Re: [PATCH v2 13/23] x86/resctrl: Allow update_mba_bw() to update controls directly

[James Morse <james.morse@arm.com>]

Hi Reinette,

On 15/10/2021 23:28, Reinette Chatre wrote:
> On 10/1/2021 9:02 AM, James Morse wrote:
>> diff --git a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>> b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>> index 9f45207a6c74..25baacd331e0 100644
>> --- a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>> +++ b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
>> @@ -282,6 +282,27 @@ static bool apply_config(struct rdt_hw_domain *hw_dom,
>>       return false;
>>   }
>>   +int resctrl_arch_update_one(struct rdt_resource *r, struct rdt_domain *d,
>> +                u32 closid, enum resctrl_conf_type t, u32 cfg_val)
>> +{
>> +    struct rdt_hw_domain *hw_dom = resctrl_to_arch_dom(d);
>> +    u32 idx = get_config_index(closid, t);
>> +    struct msr_param msr_param;
>> +
>> +    if (!cpumask_test_cpu(smp_processor_id(), &d->cpu_mask))
>> +        return -EINVAL;
>> +
>> +    hw_dom->ctrl_val[idx] = cfg_val;
>> +
>> +    msr_param.res = r;
>> +    msr_param.low = idx;
>> +    msr_param.high = idx + 1;
>> +
>> +    rdt_ctrl_update(&msr_param);
>> +
> 
> rdt_ctrl_update() will take its parameters and recompute the domain that is already
> available here ... seems to take a few steps back and then do the needed. Could msr_update
> be called directly here instead?

Even better!


Thanks,

James