From: "Deucher, Alexander" <Alexander.Deucher@amd.com>
To: "Rafael J. Wysocki" <rafael@kernel.org>,
"Su, Jinzhou (Joe)" <Jinzhou.Su@amd.com>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>,
Linux PM <linux-pm@vger.kernel.org>,
Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
Doug Smythies <dsmythies@telus.net>,
"Huang, Ray" <Ray.Huang@amd.com>,
Viresh Kumar <viresh.kumar@linaro.org>,
Todd Brandt <todd.e.brandt@linux.intel.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
"Sharma, Deepak" <Deepak.Sharma@amd.com>,
"Du, Xiaojian" <Xiaojian.Du@amd.com>,
"Yuan, Perry" <Perry.Yuan@amd.com>,
"Meng, Li (Jassmine)" <Li.Meng@amd.com>
Subject: RE: [PATCH 1/3] cpufreq: amd-pstate: Add more tracepoint for AMD P-State module
Date: Tue, 1 Mar 2022 16:05:24 +0000 [thread overview]
Message-ID: <BL1PR12MB51447B1235E91A81309190B2F7029@BL1PR12MB5144.namprd12.prod.outlook.com> (raw)
In-Reply-To: <CAJZ5v0iaY3tsNKFXv09Z4wg_2R3+9UsSnqfPBbOYFaqoVS1qCg@mail.gmail.com>
[AMD Official Use Only]
> -----Original Message-----
> From: Rafael J. Wysocki <rafael@kernel.org>
> Sent: Tuesday, March 1, 2022 10:26 AM
> To: Su, Jinzhou (Joe) <Jinzhou.Su@amd.com>
> Cc: Rafael J. Wysocki <rjw@rjwysocki.net>; Linux PM <linux-
> pm@vger.kernel.org>; Srinivas Pandruvada
> <srinivas.pandruvada@linux.intel.com>; Doug Smythies
> <dsmythies@telus.net>; Huang, Ray <Ray.Huang@amd.com>; Viresh Kumar
> <viresh.kumar@linaro.org>; Todd Brandt <todd.e.brandt@linux.intel.com>;
> Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Sharma, Deepak
> <Deepak.Sharma@amd.com>; Deucher, Alexander
> <Alexander.Deucher@amd.com>; Du, Xiaojian <Xiaojian.Du@amd.com>;
> Yuan, Perry <Perry.Yuan@amd.com>; Meng, Li (Jassmine)
> <Li.Meng@amd.com>
> Subject: Re: [PATCH 1/3] cpufreq: amd-pstate: Add more tracepoint for AMD
> P-State module
>
> On Wed, Feb 23, 2022 at 11:04 AM Jinzhou Su <Jinzhou.Su@amd.com>
> wrote:
> >
> > Add frequency, mperf, aperf and tsc in the trace. This can be used to
> > debug and tune the performance of AMD P-state driver.
> >
> > Use the time difference between amd_pstate_update to calculate CPU
> > frequency. There could be sleep in arch_freq_get_on_cpu, so do not use
> > it here.
> >
> > Signed-off-by: Jinzhou Su <Jinzhou.Su@amd.com>
> > Signed-off-by: Huang Rui <ray.huang@amd.com>
>
> I'm not sure what the second sign-off is for.
>
> If this is a maintainer's sign-off, it should be added by the maintainer himself
> and you should not add it when submitting the patch.
Both developers co-worked on the patch. Isn't that pretty standard when you rework someone else's patch?
Alex
>
> > ---
> > drivers/cpufreq/amd-pstate-trace.h | 22 ++++++++++-
> > drivers/cpufreq/amd-pstate.c | 59
> +++++++++++++++++++++++++++++-
> > 2 files changed, 78 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/cpufreq/amd-pstate-trace.h
> > b/drivers/cpufreq/amd-pstate-trace.h
> > index 647505957d4f..35f38ae67fb1 100644
> > --- a/drivers/cpufreq/amd-pstate-trace.h
> > +++ b/drivers/cpufreq/amd-pstate-trace.h
> > @@ -27,6 +27,10 @@ TRACE_EVENT(amd_pstate_perf,
> > TP_PROTO(unsigned long min_perf,
> > unsigned long target_perf,
> > unsigned long capacity,
> > + u64 freq,
> > + u64 mperf,
> > + u64 aperf,
> > + u64 tsc,
> > unsigned int cpu_id,
> > bool changed,
> > bool fast_switch
> > @@ -35,6 +39,10 @@ TRACE_EVENT(amd_pstate_perf,
> > TP_ARGS(min_perf,
> > target_perf,
> > capacity,
> > + freq,
> > + mperf,
> > + aperf,
> > + tsc,
> > cpu_id,
> > changed,
> > fast_switch
> > @@ -44,6 +52,10 @@ TRACE_EVENT(amd_pstate_perf,
> > __field(unsigned long, min_perf)
> > __field(unsigned long, target_perf)
> > __field(unsigned long, capacity)
> > + __field(unsigned long long, freq)
> > + __field(unsigned long long, mperf)
> > + __field(unsigned long long, aperf)
> > + __field(unsigned long long, tsc)
> > __field(unsigned int, cpu_id)
> > __field(bool, changed)
> > __field(bool, fast_switch) @@ -53,15 +65,23 @@
> > TRACE_EVENT(amd_pstate_perf,
> > __entry->min_perf = min_perf;
> > __entry->target_perf = target_perf;
> > __entry->capacity = capacity;
> > + __entry->freq = freq;
> > + __entry->mperf = mperf;
> > + __entry->aperf = aperf;
> > + __entry->tsc = tsc;
> > __entry->cpu_id = cpu_id;
> > __entry->changed = changed;
> > __entry->fast_switch = fast_switch;
> > ),
> >
> > - TP_printk("amd_min_perf=%lu amd_des_perf=%lu
> amd_max_perf=%lu cpu_id=%u changed=%s fast_switch=%s",
> > + TP_printk("amd_min_perf=%lu amd_des_perf=%lu
> amd_max_perf=%lu
> > + freq=%llu mperf=%llu aperf=%llu tsc=%llu cpu_id=%u changed=%s
> > + fast_switch=%s",
> > (unsigned long)__entry->min_perf,
> > (unsigned long)__entry->target_perf,
> > (unsigned long)__entry->capacity,
> > + (unsigned long long)__entry->freq,
> > + (unsigned long long)__entry->mperf,
> > + (unsigned long long)__entry->aperf,
> > + (unsigned long long)__entry->tsc,
> > (unsigned int)__entry->cpu_id,
> > (__entry->changed) ? "true" : "false",
> > (__entry->fast_switch) ? "true" : "false"
> > diff --git a/drivers/cpufreq/amd-pstate.c
> > b/drivers/cpufreq/amd-pstate.c index 9ce75ed11f8e..7be38bc6a673
> 100644
> > --- a/drivers/cpufreq/amd-pstate.c
> > +++ b/drivers/cpufreq/amd-pstate.c
> > @@ -65,6 +65,18 @@ MODULE_PARM_DESC(shared_mem,
> >
> > static struct cpufreq_driver amd_pstate_driver;
> >
> > +/**
> > + * struct amd_aperf_mperf
> > + * @aperf: actual performance frequency clock count
> > + * @mperf: maximum performance frequency clock count
> > + * @tsc: time stamp counter
> > + */
> > +struct amd_aperf_mperf {
> > + u64 aperf;
> > + u64 mperf;
> > + u64 tsc;
> > +};
> > +
> > /**
> > * struct amd_cpudata - private CPU data for AMD P-State
> > * @cpu: CPU number
> > @@ -81,6 +93,9 @@ static struct cpufreq_driver amd_pstate_driver;
> > * @min_freq: the frequency that mapped to lowest_perf
> > * @nominal_freq: the frequency that mapped to nominal_perf
> > * @lowest_nonlinear_freq: the frequency that mapped to
> > lowest_nonlinear_perf
> > + * @cur: Difference of Aperf/Mperf/tsc count between last and current
> > + sample
> > + * @prev: Last Aperf/Mperf/tsc count value read from register
> > + * @freq: current cpu frequency value
> > * @boost_supported: check whether the Processor or SBIOS supports
> boost mode
> > *
> > * The amd_cpudata is key private data for each CPU thread in AMD
> > P-State, and @@ -102,6 +117,10 @@ struct amd_cpudata {
> > u32 nominal_freq;
> > u32 lowest_nonlinear_freq;
> >
> > + struct amd_aperf_mperf cur;
> > + struct amd_aperf_mperf prev;
> > +
> > + u64 freq;
> > bool boost_supported;
> > };
> >
> > @@ -211,6 +230,39 @@ static inline void amd_pstate_update_perf(struct
> amd_cpudata *cpudata,
> > max_perf, fast_switch); }
> >
> > +static inline bool amd_pstate_sample(struct amd_cpudata *cpudata) {
> > + u64 aperf, mperf, tsc;
> > + unsigned long flags;
> > +
> > + local_irq_save(flags);
> > + rdmsrl(MSR_IA32_APERF, aperf);
> > + rdmsrl(MSR_IA32_MPERF, mperf);
> > + tsc = rdtsc();
> > +
> > + if (cpudata->prev.mperf == mperf || cpudata->prev.tsc == tsc) {
> > + local_irq_restore(flags);
> > + return false;
> > + }
> > +
> > + local_irq_restore(flags);
> > +
> > + cpudata->cur.aperf = aperf;
> > + cpudata->cur.mperf = mperf;
> > + cpudata->cur.tsc = tsc;
> > + cpudata->cur.aperf -= cpudata->prev.aperf;
> > + cpudata->cur.mperf -= cpudata->prev.mperf;
> > + cpudata->cur.tsc -= cpudata->prev.tsc;
> > +
> > + cpudata->prev.aperf = aperf;
> > + cpudata->prev.mperf = mperf;
> > + cpudata->prev.tsc = tsc;
> > +
> > + cpudata->freq = div64_u64((cpudata->cur.aperf * cpu_khz),
> > + cpudata->cur.mperf);
> > +
> > + return true;
> > +}
> > +
> > static void amd_pstate_update(struct amd_cpudata *cpudata, u32
> min_perf,
> > u32 des_perf, u32 max_perf, bool
> > fast_switch) { @@ -226,8 +278,11 @@ static void
> > amd_pstate_update(struct amd_cpudata *cpudata, u32 min_perf,
> > value &= ~AMD_CPPC_MAX_PERF(~0L);
> > value |= AMD_CPPC_MAX_PERF(max_perf);
> >
> > - trace_amd_pstate_perf(min_perf, des_perf, max_perf,
> > - cpudata->cpu, (value != prev), fast_switch);
> > + if (trace_amd_pstate_perf_enabled() &&
> amd_pstate_sample(cpudata)) {
> > + trace_amd_pstate_perf(min_perf, des_perf, max_perf, cpudata-
> >freq,
> > + cpudata->cur.mperf, cpudata->cur.aperf, cpudata->cur.tsc,
> > + cpudata->cpu, (value != prev), fast_switch);
> > + }
> >
> > if (value == prev)
> > return;
> > --
> > 2.27.0
> >
next prev parent reply other threads:[~2022-03-01 16:05 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-23 10:03 [PATCH 0/3] Add tracer tool for AMD P-State driver Jinzhou Su
2022-02-23 10:03 ` [PATCH 1/3] cpufreq: amd-pstate: Add more tracepoint for AMD P-State module Jinzhou Su
2022-03-01 15:26 ` Rafael J. Wysocki
2022-03-01 16:05 ` Deucher, Alexander [this message]
2022-03-01 16:07 ` Rafael J. Wysocki
2022-03-02 9:09 ` Huang Rui
2022-03-02 9:53 ` Su, Jinzhou (Joe)
2022-02-23 10:03 ` [PATCH 2/3] tools/power/x86/intel_pstate_tracer: make tracer as a module Jinzhou Su
2022-02-23 15:19 ` Doug Smythies
2022-02-23 10:03 ` [PATCH 3/3] tools/power/x86/amd_pstate_tracer: Add tracer tool for AMD P-state Jinzhou Su
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=BL1PR12MB51447B1235E91A81309190B2F7029@BL1PR12MB5144.namprd12.prod.outlook.com \
--to=alexander.deucher@amd.com \
--cc=Deepak.Sharma@amd.com \
--cc=Jinzhou.Su@amd.com \
--cc=Li.Meng@amd.com \
--cc=Perry.Yuan@amd.com \
--cc=Ray.Huang@amd.com \
--cc=Xiaojian.Du@amd.com \
--cc=dsmythies@telus.net \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=rafael@kernel.org \
--cc=rjw@rjwysocki.net \
--cc=srinivas.pandruvada@linux.intel.com \
--cc=todd.e.brandt@linux.intel.com \
--cc=viresh.kumar@linaro.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).