[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <cfd06af3-a950-400a-a858-1dd67b968e69@amd.com>
Date: Tue, 11 Jun 2024 19:47:09 +0530
From: Dhananjay Ugwekar <Dhananjay.Ugwekar@....com>
To: "Zhang, Rui" <rui.zhang@...el.com>,
"alexander.shishkin@...ux.intel.com" <alexander.shishkin@...ux.intel.com>,
"dave.hansen@...ux.intel.com" <dave.hansen@...ux.intel.com>,
"Hunter, Adrian" <adrian.hunter@...el.com>,
"mingo@...hat.com" <mingo@...hat.com>,
"irogers@...gle.com" <irogers@...gle.com>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"gustavoars@...nel.org" <gustavoars@...nel.org>,
"kan.liang@...ux.intel.com" <kan.liang@...ux.intel.com>,
"kees@...nel.org" <kees@...nel.org>,
"mark.rutland@....com" <mark.rutland@....com>,
"peterz@...radead.org" <peterz@...radead.org>, "bp@...en8.de"
<bp@...en8.de>, "acme@...nel.org" <acme@...nel.org>,
"jolsa@...nel.org" <jolsa@...nel.org>, "x86@...nel.org" <x86@...nel.org>,
"namhyung@...nel.org" <namhyung@...nel.org>
Cc: "ravi.bangoria@....com" <ravi.bangoria@....com>,
"kprateek.nayak@....com" <kprateek.nayak@....com>,
"gautham.shenoy@....com" <gautham.shenoy@....com>,
"linux-perf-users@...r.kernel.org" <linux-perf-users@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-hardening@...r.kernel.org" <linux-hardening@...r.kernel.org>,
"sandipan.das@....com" <sandipan.das@....com>,
"ananth.narayan@....com" <ananth.narayan@....com>,
"linux-pm@...r.kernel.org" <linux-pm@...r.kernel.org>
Subject: Re: [PATCH 1/6] perf/x86/rapl: Fix the energy-pkg event for AMD CPUs
Hello Rui,
On 6/11/2024 11:05 AM, Zhang, Rui wrote:
> On Mon, 2024-06-10 at 10:07 +0000, Dhananjay Ugwekar wrote:
>> After commit ("x86/cpu/topology: Add support for the AMD 0x80000026
>> leaf"),
>> on AMD processors that support extended CPUID leaf 0x80000026, the
>> topology_die_cpumask() and topology_logical_die_id() macros, no
>> longer
>> return the package cpumask and package id, instead they return the
>> CCD
>> (Core Complex Die) mask and id respectively. This leads to the
>> energy-pkg
>> event scope to be modified to CCD instead of package.
>>
>> Replacing these macros with their package counterparts fixes the
>> energy-pkg event for AMD CPUs.
>>
>> However due to the difference between the scope of energy-pkg event
>> for
>> Intel and AMD CPUs,
>
> Actually, on Intel platforms, the energy-pkg event is also package
> scope except one platform, and that one is the only multi-die system
> with RAPL MSRs available.
> So that is why the die/pkg logic was introduced in the first place.
>
> I like the macro/helpers in this patch. The logic inside them may need
> to be optimized for Intel platforms, but that is a separate task.
Yes that is correct, however anyway the die macros return the package
values on a single die system, so I preferred to use them for all Intel
systems (single or multi-die).
We can surely update the macro name or logic in future to make it more
semantically correct.
Thanks for the review!
Regards,
Dhananjay
>
>> we have to replace these macros conditionally only for
>> AMD CPUs.
>>
>> On a 12 CCD 1 Package AMD Zen4 Genoa machine:
>>
>> Before:
>> $ cat /sys/devices/power/cpumask
>> 0,8,16,24,32,40,48,56,64,72,80,88.
>>
>> The expected cpumask here is supposed to be just "0", as it is a
>> package
>> scope event, only one CPU will be collecting the event for all the
>> CPUs in
>> the package.
>>
>> After:
>> $ cat /sys/devices/power/cpumask
>> 0
>>
>> Signed-off-by: Dhananjay Ugwekar <Dhananjay.Ugwekar@....com>
>> Fixes: 63edbaa48a57 ("x86/cpu/topology: Add support for the AMD
>> 0x80000026 leaf")
>
> Reviewed-by: Zhang Rui <rui.zhang@...el.com>
>
> thanks,
> rui
>> ---
>> PS: This patch was earlier sent separately(link below), it has not
>> been
>> merged yet, it is necessary for this patchset to work properly, also
>> it
>> fixes the pre-existing energy-pkg event.
>> https://lore.kernel.org/linux-perf-users/20240502095115.177713-1-Dhananjay.Ugwekar@amd.com/
>> ---
>> arch/x86/events/rapl.c | 30 ++++++++++++++++++++++++++----
>> 1 file changed, 26 insertions(+), 4 deletions(-)
>>
>> diff --git a/arch/x86/events/rapl.c b/arch/x86/events/rapl.c
>> index b985ca79cf97..73be25e1f4b4 100644
>> --- a/arch/x86/events/rapl.c
>> +++ b/arch/x86/events/rapl.c
>> @@ -103,6 +103,10 @@ static struct perf_pmu_events_attr
>> event_attr_##v = { \
>> .event_str =
>> str, \
>> };
>>
>> +#define rapl_pmu_is_pkg_scope() \
>> + (boot_cpu_data.x86_vendor == X86_VENDOR_AMD || \
>> + boot_cpu_data.x86_vendor == X86_VENDOR_HYGON)
>> +
>> struct rapl_pmu {
>> raw_spinlock_t lock;
>> int n_active;
>> @@ -140,9 +144,21 @@ static unsigned int rapl_cntr_mask;
>> static u64 rapl_timer_ms;
>> static struct perf_msr *rapl_msrs;
>>
>> +static inline unsigned int get_rapl_pmu_idx(int cpu)
>> +{
>> + return rapl_pmu_is_pkg_scope() ?
>> topology_logical_package_id(cpu) :
>> +
>> topology_logical_die_id(cpu);
>> +}
>> +
>> +static inline const struct cpumask *get_rapl_pmu_cpumask(int cpu)
>> +{
>> + return rapl_pmu_is_pkg_scope() ? topology_core_cpumask(cpu) :
>> + topology_die_cpumask(cpu);
>> +}
>> +
>> static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu)
>> {
>> - unsigned int rapl_pmu_idx = topology_logical_die_id(cpu);
>> + unsigned int rapl_pmu_idx = get_rapl_pmu_idx(cpu);
>>
>> /*
>> * The unsigned check also catches the '-1' return value for
>> non
>> @@ -543,6 +559,7 @@ static struct perf_msr amd_rapl_msrs[] = {
>>
>> static int rapl_cpu_offline(unsigned int cpu)
>> {
>> + const struct cpumask *rapl_pmu_cpumask =
>> get_rapl_pmu_cpumask(cpu);
>> struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
>> int target;
>>
>> @@ -552,7 +569,7 @@ static int rapl_cpu_offline(unsigned int cpu)
>>
>> pmu->cpu = -1;
>> /* Find a new cpu to collect rapl events */
>> - target = cpumask_any_but(topology_die_cpumask(cpu), cpu);
>> + target = cpumask_any_but(rapl_pmu_cpumask, cpu);
>>
>> /* Migrate rapl events to the new target */
>> if (target < nr_cpu_ids) {
>> @@ -565,6 +582,8 @@ static int rapl_cpu_offline(unsigned int cpu)
>>
>> static int rapl_cpu_online(unsigned int cpu)
>> {
>> + unsigned int rapl_pmu_idx = get_rapl_pmu_idx(cpu);
>> + const struct cpumask *rapl_pmu_cpumask =
>> get_rapl_pmu_cpumask(cpu);
>> struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
>> int target;
>>
>> @@ -579,14 +598,14 @@ static int rapl_cpu_online(unsigned int cpu)
>> pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
>> rapl_hrtimer_init(pmu);
>>
>> - rapl_pmus->pmus[topology_logical_die_id(cpu)] = pmu;
>> + rapl_pmus->pmus[rapl_pmu_idx] = pmu;
>> }
>>
>> /*
>> * Check if there is an online cpu in the package which
>> collects rapl
>> * events already.
>> */
>> - target = cpumask_any_and(&rapl_cpu_mask,
>> topology_die_cpumask(cpu));
>> + target = cpumask_any_and(&rapl_cpu_mask, rapl_pmu_cpumask);
>> if (target < nr_cpu_ids)
>> return 0;
>>
>> @@ -677,6 +696,9 @@ static int __init init_rapl_pmus(void)
>> {
>> int nr_rapl_pmu = topology_max_packages() *
>> topology_max_dies_per_package();
>>
>> + if (rapl_pmu_is_pkg_scope())
>> + nr_rapl_pmu = topology_max_packages();
>> +
>> rapl_pmus = kzalloc(struct_size(rapl_pmus, pmus,
>> nr_rapl_pmu), GFP_KERNEL);
>> if (!rapl_pmus)
>> return -ENOMEM;
>
Powered by blists - more mailing lists