Re: [Xen-devel] [PATCH v2 6/9] x86/intel_pstate: the main boby of the intel_pstate driver

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 10.06.15 at 07:20, <wei.w.wang@xxxxxxxxx> wrote:
> On 26/05/2015 21:58, Jan Beulich wrote
>> >>> On 13.05.16 at 09:50, <wei.w.wang@xxxxxxxxx> wrote:
>> > +    if (policy->policy == CPUFREQ_POLICY_PERFORMANCE) {
>> > +        limits.no_turbo = 0;
>> > +        limits.max_perf_pct = 100;
>> > +        limits.max_perf = int_tofp(1);
>> > +        limits.min_perf_pct = 100;
>> > +        limits.min_perf = int_tofp(1);
>> > +        policy->max_perf_pct = 100;
>> > +        policy->min_perf_pct = 100;
>> > +        return 0;
> 
> I noticed another issue. The intel_pstate driver originally manages all the 
> CPUs using the global "limit" structure (e.g. if one CPU is set to be in the 
> Performance mode, min_perf=max_perf , then all the CPUs will run in the 
> Performance mode). This will generate confusing status info to users in our 
> case. For example, if we set CPU0 in Powersave, then set CPU1 in Performance. 
> We will see CPU0 in the Powersave mode, but it's actually running in the 
> Performance mode.
> 
> How do you think if we move the global "limit" structure to the per-cpu 
> policy structure, so that each CPU can be configured individually?

Sounds reasonable; I don't recall having asked for that structure to
be global...

Jan


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