Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On Mon, Jul 10, 2017 at 10:49 AM, Vikram Mulukutla wrote: [..] > >>> Given that the utilization update hooks are called with the per-cpu rq >>> lock >>> held (for all classes), I don't think PELT utilization can change >>> throughout >>> the lifetime of the

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On Mon, Jul 10, 2017 at 10:49 AM, Vikram Mulukutla wrote: [..] > >>> Given that the utilization update hooks are called with the per-cpu rq >>> lock >>> held (for all classes), I don't think PELT utilization can change >>> throughout >>> the lifetime of the cpufreq_update_{util,this_cpu} call?

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On 2017-07-07 23:14, Joel Fernandes wrote: Hi Vikram, On Thu, Jul 6, 2017 at 11:44 PM, Vikram Mulukutla wrote: On 2017-07-04 10:34, Patrick Bellasi wrote: Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On 2017-07-07 23:14, Joel Fernandes wrote: Hi Vikram, On Thu, Jul 6, 2017 at 11:44 PM, Vikram Mulukutla wrote: On 2017-07-04 10:34, Patrick Bellasi wrote: Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be a big issue for most

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

Hi Vikram, On Thu, Jul 6, 2017 at 11:44 PM, Vikram Mulukutla wrote: > On 2017-07-04 10:34, Patrick Bellasi wrote: >> >> Currently the utilization of the FAIR class is collected before locking >> the policy. Although that should not be a big issue for most cases, we >>

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

Hi Vikram, On Thu, Jul 6, 2017 at 11:44 PM, Vikram Mulukutla wrote: > On 2017-07-04 10:34, Patrick Bellasi wrote: >> >> Currently the utilization of the FAIR class is collected before locking >> the policy. Although that should not be a big issue for most cases, we >> also don't really know how

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On 2017-07-04 10:34, Patrick Bellasi wrote: Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be a big issue for most cases, we also don't really know how much latency there can be between the utilization reading and its usage. Let's

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On 2017-07-04 10:34, Patrick Bellasi wrote: Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be a big issue for most cases, we also don't really know how much latency there can be between the utilization reading and its usage. Let's

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On Tue, Jul 4, 2017 at 10:34 AM, Patrick Bellasi wrote: > Currently the utilization of the FAIR class is collected before locking > the policy. Although that should not be a big issue for most cases, we > also don't really know how much latency there can be between the >

Re: [PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

On Tue, Jul 4, 2017 at 10:34 AM, Patrick Bellasi wrote: > Currently the utilization of the FAIR class is collected before locking > the policy. Although that should not be a big issue for most cases, we > also don't really know how much latency there can be between the > utilization reading and

[PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be a big issue for most cases, we also don't really know how much latency there can be between the utilization reading and its usage. Let's get the FAIR utilization right before its usage

[PATCH v2 4/6] cpufreq: schedutil: update CFS util only if used

Currently the utilization of the FAIR class is collected before locking the policy. Although that should not be a big issue for most cases, we also don't really know how much latency there can be between the utilization reading and its usage. Let's get the FAIR utilization right before its usage