Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-10 Thread Pan, Harry
My apology that I resent again in order to clean my junk of useless enum, and ignore ambiguous in-reply-to (2016/9/11/8) due to subject. new patchset is here: https://lkml.org/lkml/2016/9/11/11 https://lkml.org/lkml/2016/9/11/12 https://lkml.org/lkml/2016/9/11/10 Thanks, Harry

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-10 Thread Pan, Harry
Hi Thomas, Appreciate comments, I understood and learned. I just uploaded 3 patches integrated yours, yet there was 'git am' failure thus I did hand work, kindly double check. Sincerely, Harry On Fri, 2016-09-09 at 23:21 +0200, Thomas Gleixner wrote: > On Fri, 9 Sep 2016, Pan, Harry wrote: > > O

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Thomas Gleixner
On Fri, 9 Sep 2016, Pan, Harry wrote: > On Fri, 2016-09-09 at 17:11 +0200, Thomas Gleixner wrote: > > > struct intel_rapl_init_fun { > > > - bool apply_quirk; > > > + enum rapl_quirk apply_quirk; > > > > This is silly. Make apply_quirk a function pointer and provide functions > > for the differen

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread kbuild test robot
Hi Harry, [auto build test WARNING on v4.8-rc5] [also build test WARNING on next-20160909] [cannot apply to tip/x86/core] [if your patch is applied to the wrong git tree, please drop us a note to help improve the system] [Suggest to use git(>=2.9.0) format-patch --base= (or --base=auto for conve

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Pan, Harry
I refined/uploaded again, kindly advise. On Fri, 2016-09-09 at 17:11 +0200, Thomas Gleixner wrote: > On Fri, 9 Sep 2016, Harry Pan wrote: > > - if (apply_quirk) > > + if (apply_quirk == RAPL_HSX_QUIRK) > > rapl_hw_unit[RAPL_IDX_RAM_NRG_STAT] = 16; > > > > /* > > +* Some A

[PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Harry Pan
This patch enables RAPL counters (energy consumption counters) support for Intel Baytrail and Braswell processors (Model 55 and 76): The Silvermont/Airmont microarchitecture actually uses fixed energy status unit (ESU) in smallest unit of microjoule, this patch adds quirk for these Atom processors

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Thomas Gleixner
On Fri, 9 Sep 2016, Harry Pan wrote: > - if (apply_quirk) > + if (apply_quirk == RAPL_HSX_QUIRK) > rapl_hw_unit[RAPL_IDX_RAM_NRG_STAT] = 16; > > /* > + * Some Atom processors (BYT/BSW) have 2^ESU microjoules increment, > + * refer to Software Developers' Manu

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Pan, Harry
Hi Peter, Totally agreed and uploaded patchset again. https://lkml.org/lkml/2016/9/9/467 https://lkml.org/lkml/2016/9/9/468 One more thing, I did not refine rapl_advertise() description. Advice is welcome. Sincerely, Harry On Fri, 2016-09-09 at 11:29 +0200, Peter Zijlstra wrote: > On Thu, Sep 08

[PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Harry Pan
This patch enables RAPL counters (energy consumption counters) support for Intel Baytrail and Braswell processors (Model 55 and 76): The Silvermont/Airmont microarchitecture actually uses fixed energy status unit (ESU) in smallest unit of microjoule, this patch adds quirk for these Atom processors

Re: [PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-09 Thread Peter Zijlstra
On Thu, Sep 08, 2016 at 05:08:58PM +0800, Harry Pan wrote: > @@ -177,6 +187,16 @@ static inline u64 rapl_scale(u64 v, int cfg) > pr_warn("Invalid domain %d, failed to scale data\n", cfg); > return v; > } > + > + /* > + * Some Atom series processors (BYT/B

[PATCH 2/2] perf/x86/rapl: Enable Baytrail/Braswell RAPL support

2016-09-08 Thread Harry Pan
This patch enables RAPL counters (energy consumption counters) support for Intel Baytrail and Braswell processors (Model 55 and 76): The Silvermont/Airmont microarchitecture actually uses fixed energy status unit (ESU) in smallest unit of microjoule, this patch adds quirk for these Atom processors