From: "Liang, Kan" <kan.liang@linux.intel.com>
To: Ian Rogers <irogers@google.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>,
Mike Leach <mike.leach@linaro.org>, Leo Yan <leo.yan@linaro.org>,
John Garry <john.g.garry@oracle.com>,
Will Deacon <will@kernel.org>, James Clark <james.clark@arm.com>,
Peter Zijlstra <peterz@infradead.org>,
Ingo Molnar <mingo@redhat.com>,
Arnaldo Carvalho de Melo <acme@kernel.org>,
Mark Rutland <mark.rutland@arm.com>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Jiri Olsa <jolsa@kernel.org>, Namhyung Kim <namhyung@kernel.org>,
Adrian Hunter <adrian.hunter@intel.com>,
Kajol Jain <kjain@linux.ibm.com>,
Jing Zhang <renyu.zj@linux.alibaba.com>,
Ravi Bangoria <ravi.bangoria@amd.com>,
Madhavan Srinivasan <maddy@linux.ibm.com>,
Athira Rajeev <atrajeev@linux.vnet.ibm.com>,
Ming Wang <wangming01@loongson.cn>,
Huacai Chen <chenhuacai@kernel.org>,
Sandipan Das <sandipan.das@amd.com>,
Dmitrii Dolgov <9erthalion6@gmail.com>,
Sean Christopherson <seanjc@google.com>,
Ali Saidi <alisaidi@amazon.com>, Rob Herring <robh@kernel.org>,
Thomas Richter <tmricht@linux.ibm.com>,
Kang Minchul <tegongkang@gmail.com>,
linux-kernel@vger.kernel.org, coresight@lists.linaro.org,
linux-arm-kernel@lists.infradead.org,
linux-perf-users@vger.kernel.org
Subject: Re: [PATCH v1 05/23] perf pmu: Remove perf_pmu__hybrid_mounted
Date: Mon, 22 May 2023 07:55:27 -0400 [thread overview]
Message-ID: <642c0b26-fcd8-8063-72d2-c74466c29785@linux.intel.com> (raw)
In-Reply-To: <CAP-5=fU3OYNsjaJhLq_8tJXZKFARCTqrg3CEr64-3rAm_kTx+A@mail.gmail.com>
On 2023-05-22 1:21 a.m., Ian Rogers wrote:
> On Sun, May 21, 2023 at 12:23 PM Liang, Kan <kan.liang@linux.intel.com> wrote:
>>
>>
>>
>> On 2023-05-17 10:57 a.m., Ian Rogers wrote:
>>> perf_pmu__hybrid_mounted is used to detect whether cpu_core or
>>> cpu_atom
>>
>> Currently, there are only two CPU types for a hybrid machine, core and
>> atom. But there may be more CPU types added later. Please see the CPUID
>> 1AH EAX enumeration in SDM VOL2. It has several reserved encodings for
>> CPU types. It's better not using the hardcode cpu_core/cpu_atom to
>> replace the perf_pmu__hybrid_mounted().
>>
>> Thanks,
>> Kan
>
> This is covered by later patches. Specifically patches like:
> patch 7: perf pmu: Add is_core to pmu
> patch 20: Split pmus list into core and uncore
>
> Ultimately in pmus.h we have two scan routines, one for all PMUs and
> one for core PMUs. For everything except hybrid (and the name varies a
> bit on ARM) the core scan only scans "cpu", on hybrid it scans
> "cpu_atom" and "cpu_core". The determination of core vs uncore is
> done without using the name, so can support >2 hybrid PMUs. At this
> point in the patch series I'm trying to simplify everything so that I
> can then build the pmus interface.
But if we add a new core type "cpu_whatever" later, we have to hardcode
the new name to the perf tool, right? Users have to update the perf tool
for the new platforms, otherwise I think the new type will be treated as
an uncore PMU.
Since the hybrid is Intel only, I think it may be better move the
is_pmu_hybrid() to X86 specifc code. For the Intel only code, we already
have a naming rule for the hybrid name, "cpu_$". So we don't need to
update the tool for every new CPU type.
Thanks,
Kan
>
> Thanks,
> Ian
>
>>> is mounted with a non-empty cpus file by
>>> pmu_lookup. pmu_lookup will attempt to read the cpus file too and so
>>> the check can be folded into this.
>>>
>>> Checking hybrid_mounted in pmu_is_uncore is redundant as the next
>>> cpumask read will fail returning false.
>>>
>>> Reduce the scope of perf_pmu__find_hybrid_pmu by making it static.
>>>
>>> Signed-off-by: Ian Rogers <irogers@google.com>
>>> ---
>>> tools/perf/util/pmu-hybrid.c | 15 +--------------
>>> tools/perf/util/pmu-hybrid.h | 3 ---
>>> tools/perf/util/pmu.c | 26 ++++++++++++++------------
>>> 3 files changed, 15 insertions(+), 29 deletions(-)
>>>
>>> diff --git a/tools/perf/util/pmu-hybrid.c b/tools/perf/util/pmu-hybrid.c
>>> index bc4cb0738c35..7fe943dd3217 100644
>>> --- a/tools/perf/util/pmu-hybrid.c
>>> +++ b/tools/perf/util/pmu-hybrid.c
>>> @@ -18,20 +18,7 @@
>>>
>>> LIST_HEAD(perf_pmu__hybrid_pmus);
>>>
>>> -bool perf_pmu__hybrid_mounted(const char *name)
>>> -{
>>> - int cpu;
>>> - char pmu_name[PATH_MAX];
>>> - struct perf_pmu pmu = {.name = pmu_name};
>>> -
>>> - if (strncmp(name, "cpu_", 4))
>>> - return false;
>>> -
>>> - strlcpy(pmu_name, name, sizeof(pmu_name));
>>> - return perf_pmu__scan_file(&pmu, "cpus", "%u", &cpu) > 0;
>>> -}
>>> -
>>> -struct perf_pmu *perf_pmu__find_hybrid_pmu(const char *name)
>>> +static struct perf_pmu *perf_pmu__find_hybrid_pmu(const char *name)
>>> {
>>> struct perf_pmu *pmu;
>>>
>>> diff --git a/tools/perf/util/pmu-hybrid.h b/tools/perf/util/pmu-hybrid.h
>>> index 206b94931531..8dbcae935020 100644
>>> --- a/tools/perf/util/pmu-hybrid.h
>>> +++ b/tools/perf/util/pmu-hybrid.h
>>> @@ -13,9 +13,6 @@ extern struct list_head perf_pmu__hybrid_pmus;
>>> #define perf_pmu__for_each_hybrid_pmu(pmu) \
>>> list_for_each_entry(pmu, &perf_pmu__hybrid_pmus, hybrid_list)
>>>
>>> -bool perf_pmu__hybrid_mounted(const char *name);
>>> -
>>> -struct perf_pmu *perf_pmu__find_hybrid_pmu(const char *name);
>>> bool perf_pmu__is_hybrid(const char *name);
>>>
>>> static inline int perf_pmu__hybrid_pmu_num(void)
>>> diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
>>> index 1e0be23d4dd7..729b1f166f80 100644
>>> --- a/tools/perf/util/pmu.c
>>> +++ b/tools/perf/util/pmu.c
>>> @@ -617,9 +617,6 @@ static bool pmu_is_uncore(int dirfd, const char *name)
>>> {
>>> int fd;
>>>
>>> - if (perf_pmu__hybrid_mounted(name))
>>> - return false;
>>> -
>>> fd = perf_pmu__pathname_fd(dirfd, name, "cpumask", O_PATH);
>>> if (fd < 0)
>>> return false;
>>> @@ -898,6 +895,16 @@ static int pmu_max_precise(int dirfd, struct perf_pmu *pmu)
>>> return max_precise;
>>> }
>>>
>>> +/**
>>> + * perf_pmu__skip_empty_cpus() - should pmu_lookup skip the named PMU if the
>>> + * cpus or cpumask file isn't present?
>>> + * @name: Name of PMU.
>>> + */
>>> +static bool perf_pmu__skip_empty_cpus(const char *name)
>>> +{
>>> + return !strcmp(name, "cpu_core") || !strcmp(name, "cpu_atom");
>>> +}
>>> +
>>> static struct perf_pmu *pmu_lookup(int dirfd, const char *lookup_name)
>>> {
>>> struct perf_pmu *pmu;
>>> @@ -905,15 +912,8 @@ static struct perf_pmu *pmu_lookup(int dirfd, const char *lookup_name)
>>> LIST_HEAD(aliases);
>>> __u32 type;
>>> char *name = pmu_find_real_name(lookup_name);
>>> - bool is_hybrid = perf_pmu__hybrid_mounted(name);
>>> char *alias_name;
>>>
>>> - /*
>>> - * Check pmu name for hybrid and the pmu may be invalid in sysfs
>>> - */
>>> - if (!strncmp(name, "cpu_", 4) && !is_hybrid)
>>> - return NULL;
>>> -
>>> /*
>>> * The pmu data we store & need consists of the pmu
>>> * type value and format definitions. Load both right
>>> @@ -933,8 +933,10 @@ static struct perf_pmu *pmu_lookup(int dirfd, const char *lookup_name)
>>> return NULL;
>>>
>>> pmu->cpus = pmu_cpumask(dirfd, name);
>>> - pmu->name = strdup(name);
>>> + if (!pmu->cpus && perf_pmu__skip_empty_cpus(name))
>>> + goto err;
>>>
>>> + pmu->name = strdup(name);
>>> if (!pmu->name)
>>> goto err;
>>>
>>> @@ -964,7 +966,7 @@ static struct perf_pmu *pmu_lookup(int dirfd, const char *lookup_name)
>>> list_splice(&aliases, &pmu->aliases);
>>> list_add_tail(&pmu->list, &pmus);
>>>
>>> - if (is_hybrid)
>>> + if (!strcmp(name, "cpu_core") || !strcmp(name, "cpu_atom"))
>>> list_add_tail(&pmu->hybrid_list, &perf_pmu__hybrid_pmus);
>>> else
>>> INIT_LIST_HEAD(&pmu->hybrid_list);
next prev parent reply other threads:[~2023-05-22 11:55 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-17 14:57 [PATCH v1 00/23] PMU refactoring and improvements Ian Rogers
2023-05-17 14:57 ` [PATCH v1 01/23] perf tools: Warn if no user requested CPUs match PMU's CPUs Ian Rogers
2023-05-21 19:04 ` Liang, Kan
2023-05-22 4:52 ` Ian Rogers
2023-05-22 11:33 ` Liang, Kan
2023-05-17 14:57 ` [PATCH v1 02/23] perf evlist: Remove evlist__warn_hybrid_group Ian Rogers
2023-05-17 14:57 ` [PATCH v1 03/23] perf evlist: Remove __evlist__add_default Ian Rogers
2023-05-21 19:12 ` Liang, Kan
2023-05-22 5:12 ` Ian Rogers
2023-05-17 14:57 ` [PATCH v1 04/23] perf evlist: Reduce scope of evlist__has_hybrid Ian Rogers
2023-05-17 14:57 ` [PATCH v1 05/23] perf pmu: Remove perf_pmu__hybrid_mounted Ian Rogers
2023-05-21 19:23 ` Liang, Kan
2023-05-22 5:21 ` Ian Rogers
2023-05-22 11:55 ` Liang, Kan [this message]
2023-05-22 14:06 ` Ian Rogers
2023-05-23 17:23 ` Liang, Kan
2023-05-23 17:45 ` Ian Rogers
2023-05-17 14:57 ` [PATCH v1 06/23] perf pmu: Detect ARM and hybrid PMUs with sysfs Ian Rogers
2023-05-17 14:57 ` [PATCH v1 07/23] perf pmu: Add is_core to pmu Ian Rogers
2023-05-17 14:57 ` [PATCH v1 08/23] perf pmu: Rewrite perf_pmu__has_hybrid to avoid list Ian Rogers
2023-05-17 14:57 ` [PATCH v1 09/23] perf x86: Iterate hybrid PMUs as core PMUs Ian Rogers
2023-05-17 14:57 ` [PATCH v1 10/23] perf topology: Avoid hybrid list for hybrid topology Ian Rogers
2023-05-17 14:57 ` [PATCH v1 11/23] perf evsel: Compute is_hybrid from PMU being core Ian Rogers
2023-05-17 14:57 ` [PATCH v1 12/23] perf header: Avoid hybrid PMU list in write_pmu_caps Ian Rogers
2023-05-17 14:57 ` [PATCH v1 13/23] perf metrics: Remove perf_pmu__is_hybrid use Ian Rogers
2023-05-17 14:57 ` [PATCH v1 14/23] perf stat: Avoid hybrid PMU list Ian Rogers
2023-05-17 14:57 ` [PATCH v1 15/23] perf mem: " Ian Rogers
2023-05-17 14:57 ` [PATCH v1 16/23] perf pmu: Remove perf_pmu__hybrid_pmus list Ian Rogers
2023-05-17 14:57 ` [PATCH v1 17/23] perf pmus: Prefer perf_pmu__scan over perf_pmus__for_each_pmu Ian Rogers
2023-05-21 19:43 ` Liang, Kan
2023-05-17 14:57 ` [PATCH v1 18/23] perf x86 mem: minor refactor to is_mem_loads_aux_event Ian Rogers
2023-05-21 19:47 ` Liang, Kan
2023-05-17 14:57 ` [PATCH v1 19/23] perf pmu: Separate pmu and pmus Ian Rogers
2023-05-17 14:58 ` [PATCH v1 20/23] perf pmus: Split pmus list into core and uncore Ian Rogers
2023-05-21 20:02 ` Liang, Kan
2023-05-22 5:30 ` Ian Rogers
2023-05-17 14:58 ` [PATCH v1 21/23] perf pmus: Allow just core PMU scanning Ian Rogers
2023-05-17 14:58 ` [PATCH v1 22/23] perf pmus: Avoid repeated sysfs scanning Ian Rogers
2023-05-17 14:58 ` [PATCH v1 23/23] perf pmus: Ensure all PMUs are read for find_by_type Ian Rogers
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=642c0b26-fcd8-8063-72d2-c74466c29785@linux.intel.com \
--to=kan.liang@linux.intel.com \
--cc=9erthalion6@gmail.com \
--cc=acme@kernel.org \
--cc=adrian.hunter@intel.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=alisaidi@amazon.com \
--cc=atrajeev@linux.vnet.ibm.com \
--cc=chenhuacai@kernel.org \
--cc=coresight@lists.linaro.org \
--cc=irogers@google.com \
--cc=james.clark@arm.com \
--cc=john.g.garry@oracle.com \
--cc=jolsa@kernel.org \
--cc=kjain@linux.ibm.com \
--cc=leo.yan@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=maddy@linux.ibm.com \
--cc=mark.rutland@arm.com \
--cc=mike.leach@linaro.org \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=peterz@infradead.org \
--cc=ravi.bangoria@amd.com \
--cc=renyu.zj@linux.alibaba.com \
--cc=robh@kernel.org \
--cc=sandipan.das@amd.com \
--cc=seanjc@google.com \
--cc=suzuki.poulose@arm.com \
--cc=tegongkang@gmail.com \
--cc=tmricht@linux.ibm.com \
--cc=wangming01@loongson.cn \
--cc=will@kernel.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).