Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-21 Thread Joel Fernandes
On Thu, Jul 20, 2017 at 9:09 PM, Viresh Kumar wrote: > On 20-07-17, 12:49, Joel Fernandes wrote: >> Yes I think that's fine, I thought about it some more and I think this >> can be an issue in a scenario where >> >> iowait_boost_max < policy->min but: Uhh I meant to say

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-21 Thread Joel Fernandes
On Thu, Jul 20, 2017 at 9:09 PM, Viresh Kumar wrote: > On 20-07-17, 12:49, Joel Fernandes wrote: >> Yes I think that's fine, I thought about it some more and I think this >> can be an issue in a scenario where >> >> iowait_boost_max < policy->min but: Uhh I meant to say here iowait_boost <

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-20 Thread Viresh Kumar
On 20-07-17, 12:49, Joel Fernandes wrote: > Yes I think that's fine, I thought about it some more and I think this > can be an issue in a scenario where > > iowait_boost_max < policy->min but: We will never have this case as boost-max is set to cpuinfo.max_freq. -- viresh

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-20 Thread Viresh Kumar
On 20-07-17, 12:49, Joel Fernandes wrote: > Yes I think that's fine, I thought about it some more and I think this > can be an issue in a scenario where > > iowait_boost_max < policy->min but: We will never have this case as boost-max is set to cpuinfo.max_freq. -- viresh

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-20 Thread Joel Fernandes
Hi Viresh, On Wed, Jul 19, 2017 at 8:41 PM, Viresh Kumar wrote: > On 19-07-17, 19:38, Joel Fernandes wrote: >> On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar >> wrote: >> > On 18-07-17, 21:39, Joel Fernandes wrote: >> >> Not really, to me B

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-20 Thread Joel Fernandes
Hi Viresh, On Wed, Jul 19, 2017 at 8:41 PM, Viresh Kumar wrote: > On 19-07-17, 19:38, Joel Fernandes wrote: >> On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar >> wrote: >> > On 18-07-17, 21:39, Joel Fernandes wrote: >> >> Not really, to me B will still work because in the case the flag is >> >>

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Viresh Kumar
On 19-07-17, 19:38, Joel Fernandes wrote: > On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar > wrote: > > On 18-07-17, 21:39, Joel Fernandes wrote: > >> Not really, to me B will still work because in the case the flag is > >> set, we are correctly double boosting in the

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Viresh Kumar
On 19-07-17, 19:38, Joel Fernandes wrote: > On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar > wrote: > > On 18-07-17, 21:39, Joel Fernandes wrote: > >> Not really, to me B will still work because in the case the flag is > >> set, we are correctly double boosting in the next cycle. > >> > >>

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Joel Fernandes
Hi Viresh, On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar wrote: > On 18-07-17, 21:39, Joel Fernandes wrote: >> Not really, to me B will still work because in the case the flag is >> set, we are correctly double boosting in the next cycle. >> >> Taking an example, with B

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Joel Fernandes
Hi Viresh, On Tue, Jul 18, 2017 at 11:19 PM, Viresh Kumar wrote: > On 18-07-17, 21:39, Joel Fernandes wrote: >> Not really, to me B will still work because in the case the flag is >> set, we are correctly double boosting in the next cycle. >> >> Taking an example, with B = flag is set and D =

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Viresh Kumar
On 18-07-17, 21:39, Joel Fernandes wrote: > Not really, to me B will still work because in the case the flag is > set, we are correctly double boosting in the next cycle. > > Taking an example, with B = flag is set and D = flag is not set > > F = Fmin (minimum) > > iowait flag B BB

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Viresh Kumar
On 18-07-17, 21:39, Joel Fernandes wrote: > Not really, to me B will still work because in the case the flag is > set, we are correctly double boosting in the next cycle. > > Taking an example, with B = flag is set and D = flag is not set > > F = Fmin (minimum) > > iowait flag B BB

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Joel Fernandes
On Tue, Jul 18, 2017 at 10:37 PM, Viresh Kumar wrote: > On 18-07-17, 15:02, Juri Lelli wrote: >> Mmm, seems to make sense to me. :/ >> >> Would the following work (on top of Joel's v5)? Rationale being that >> only in sugov_set_iowait_boost we might bump freq up (if no

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-19 Thread Joel Fernandes
On Tue, Jul 18, 2017 at 10:37 PM, Viresh Kumar wrote: > On 18-07-17, 15:02, Juri Lelli wrote: >> Mmm, seems to make sense to me. :/ >> >> Would the following work (on top of Joel's v5)? Rationale being that >> only in sugov_set_iowait_boost we might bump freq up (if no iowait_boost >> was set) or

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Viresh Kumar
On 18-07-17, 15:02, Juri Lelli wrote: > Mmm, seems to make sense to me. :/ > > Would the following work (on top of Joel's v5)? Rationale being that > only in sugov_set_iowait_boost we might bump freq up (if no iowait_boost > was set) or start from policy->min. In sugov_iowait_boost (consumer) >

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Viresh Kumar
On 18-07-17, 15:02, Juri Lelli wrote: > Mmm, seems to make sense to me. :/ > > Would the following work (on top of Joel's v5)? Rationale being that > only in sugov_set_iowait_boost we might bump freq up (if no iowait_boost > was set) or start from policy->min. In sugov_iowait_boost (consumer) >

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Joel Fernandes
Hi Viresh, I appreciate the discussion. On Mon, Jul 17, 2017 at 10:45 PM, Viresh Kumar wrote: > On 17-07-17, 10:35, Joel Fernandes wrote: >> On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar >> wrote: >> > On 16-07-17, 01:04, Joel Fernandes wrote:

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Joel Fernandes
Hi Viresh, I appreciate the discussion. On Mon, Jul 17, 2017 at 10:45 PM, Viresh Kumar wrote: > On 17-07-17, 10:35, Joel Fernandes wrote: >> On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar >> wrote: >> > On 16-07-17, 01:04, Joel Fernandes wrote: > >> >> + if (sg_cpu->iowait_boost_pending) {

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Juri Lelli
Hi, On 18/07/17 11:15, Viresh Kumar wrote: > On 17-07-17, 10:35, Joel Fernandes wrote: > > On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar > > wrote: > > > On 16-07-17, 01:04, Joel Fernandes wrote: > > > >> + if (sg_cpu->iowait_boost_pending) { > > >> +

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-18 Thread Juri Lelli
Hi, On 18/07/17 11:15, Viresh Kumar wrote: > On 17-07-17, 10:35, Joel Fernandes wrote: > > On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar > > wrote: > > > On 16-07-17, 01:04, Joel Fernandes wrote: > > > >> + if (sg_cpu->iowait_boost_pending) { > > >> +

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Viresh Kumar
On 17-07-17, 10:35, Joel Fernandes wrote: > On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar wrote: > > On 16-07-17, 01:04, Joel Fernandes wrote: > >> + if (sg_cpu->iowait_boost_pending) { > >> + sg_cpu->iowait_boost_pending = false; > >> +

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Viresh Kumar
On 17-07-17, 10:35, Joel Fernandes wrote: > On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar wrote: > > On 16-07-17, 01:04, Joel Fernandes wrote: > >> + if (sg_cpu->iowait_boost_pending) { > >> + sg_cpu->iowait_boost_pending = false; > >> + sg_cpu->iowait_boost =

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Joel Fernandes
Hi Viresh, On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar wrote: > On 16-07-17, 01:04, Joel Fernandes wrote: >> Currently the iowait_boost feature in schedutil makes the frequency go to max >> on iowait wakeups. This feature was added to handle a case that Peter >>

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Joel Fernandes
Hi Viresh, On Mon, Jul 17, 2017 at 1:04 AM, Viresh Kumar wrote: > On 16-07-17, 01:04, Joel Fernandes wrote: >> Currently the iowait_boost feature in schedutil makes the frequency go to max >> on iowait wakeups. This feature was added to handle a case that Peter >> described where the throughput

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Viresh Kumar
On 16-07-17, 01:04, Joel Fernandes wrote: > Currently the iowait_boost feature in schedutil makes the frequency go to max > on iowait wakeups. This feature was added to handle a case that Peter > described where the throughput of operations involving continuous I/O requests > [1] is reduced due

Re: [PATCH RFC v5] cpufreq: schedutil: Make iowait boost more energy efficient

2017-07-17 Thread Viresh Kumar
On 16-07-17, 01:04, Joel Fernandes wrote: > Currently the iowait_boost feature in schedutil makes the frequency go to max > on iowait wakeups. This feature was added to handle a case that Peter > described where the throughput of operations involving continuous I/O requests > [1] is reduced due