linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Namhyung Kim <namhyung@kernel.org>
To: Arnaldo Carvalho de Melo <acme@kernel.org>, Jiri Olsa <jolsa@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Ian Rogers <irogers@google.com>,
	Adrian Hunter <adrian.hunter@intel.com>,
	linux-perf-users@vger.kernel.org,
	Kan Liang <kan.liang@linux.intel.com>,
	Leo Yan <leo.yan@linaro.org>, Andi Kleen <ak@linux.intel.com>,
	Athira Rajeev <atrajeev@linux.vnet.ibm.com>,
	James Clark <james.clark@arm.com>,
	Xing Zhengjun <zhengjun.xing@linux.intel.com>
Subject: [PATCH 14/19] perf stat: Add perf_stat_merge_counters()
Date: Sun,  9 Oct 2022 22:35:55 -0700	[thread overview]
Message-ID: <20221010053600.272854-15-namhyung@kernel.org> (raw)
In-Reply-To: <20221010053600.272854-1-namhyung@kernel.org>

The perf_stat_merge_counters() is to aggregate the same events in different
PMUs like in case of uncore or hybrid.  The same logic is in the stat-display
routines but I think it should be handled when it processes the event counters.

As it works on the aggr_counters, it doesn't change the output yet.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/builtin-stat.c |  2 +
 tools/perf/util/stat.c    | 96 +++++++++++++++++++++++++++++++++++++++
 tools/perf/util/stat.h    |  2 +
 3 files changed, 100 insertions(+)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 49a7e290d778..f90e8f29cb23 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -486,6 +486,8 @@ static void process_counters(void)
 			pr_warning("failed to process counter %s\n", counter->name);
 		counter->err = 0;
 	}
+
+	perf_stat_merge_counters(&stat_config, evsel_list);
 }
 
 static void process_interval(void)
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
index 4edfc1c5dc07..1bb197782a34 100644
--- a/tools/perf/util/stat.c
+++ b/tools/perf/util/stat.c
@@ -575,6 +575,102 @@ int perf_stat_process_counter(struct perf_stat_config *config,
 	return 0;
 }
 
+static int evsel__merge_aggr_counters(struct evsel *evsel, struct evsel *alias)
+{
+	struct perf_stat_evsel *ps_a = evsel->stats;
+	struct perf_stat_evsel *ps_b = alias->stats;
+	int i;
+
+	if (ps_a->aggr == NULL && ps_b->aggr == NULL)
+		return 0;
+
+	if (ps_a->nr_aggr != ps_b->nr_aggr) {
+		pr_err("Unmatched aggregation mode between aliases\n");
+		return -1;
+	}
+
+	for (i = 0; i < ps_a->nr_aggr; i++) {
+		struct perf_counts_values *aggr_counts_a = &ps_a->aggr[i].counts;
+		struct perf_counts_values *aggr_counts_b = &ps_b->aggr[i].counts;
+
+		/* NB: don't increase aggr.nr for aliases */
+
+		aggr_counts_a->val += aggr_counts_b->val;
+		aggr_counts_a->ena += aggr_counts_b->ena;
+		aggr_counts_a->run += aggr_counts_b->run;
+	}
+
+	return 0;
+}
+/* events should have the same name, scale, unit, cgroup but on different PMUs */
+static bool evsel__is_alias(struct evsel *evsel_a, struct evsel *evsel_b)
+{
+	if (strcmp(evsel__name(evsel_a), evsel__name(evsel_b)))
+		return false;
+
+	if (evsel_a->scale != evsel_b->scale)
+		return false;
+
+	if (evsel_a->cgrp != evsel_b->cgrp)
+		return false;
+
+	if (strcmp(evsel_a->unit, evsel_b->unit))
+		return false;
+
+	if (evsel__is_clock(evsel_a) != evsel__is_clock(evsel_b))
+		return false;
+
+	return !!strcmp(evsel_a->pmu_name, evsel_b->pmu_name);
+}
+
+static void evsel__merge_aliases(struct evsel *evsel)
+{
+	struct evlist *evlist = evsel->evlist;
+	struct evsel *alias;
+
+	alias = list_prepare_entry(evsel, &(evlist->core.entries), core.node);
+	list_for_each_entry_continue(alias, &evlist->core.entries, core.node) {
+		/* Merge the same events on different PMUs. */
+		if (evsel__is_alias(evsel, alias)) {
+			evsel__merge_aggr_counters(evsel, alias);
+			alias->merged_stat = true;
+		}
+	}
+}
+
+static bool evsel__should_merge_hybrid(struct evsel *evsel, struct perf_stat_config *config)
+{
+	struct perf_pmu *pmu;
+
+	if (!config->hybrid_merge)
+		return false;
+
+	pmu = evsel__find_pmu(evsel);
+	return pmu && pmu->is_hybrid;
+}
+
+static void evsel__merge_stats(struct evsel *evsel, struct perf_stat_config *config)
+{
+	/* this evsel is already merged */
+	if (evsel->merged_stat)
+		return;
+
+	if (evsel->auto_merge_stats || evsel__should_merge_hybrid(evsel, config))
+		evsel__merge_aliases(evsel);
+}
+
+/* merge the same uncore and hybrid events if requested */
+void perf_stat_merge_counters(struct perf_stat_config *config, struct evlist *evlist)
+{
+	struct evsel *evsel;
+
+	if (config->no_merge)
+		return;
+
+	evlist__for_each_entry(evlist, evsel)
+		evsel__merge_stats(evsel, config);
+}
+
 int perf_event__process_stat_event(struct perf_session *session,
 				   union perf_event *event)
 {
diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h
index 3a876ad2870b..12cc60ab04e4 100644
--- a/tools/perf/util/stat.h
+++ b/tools/perf/util/stat.h
@@ -270,6 +270,8 @@ void evlist__reset_aggr_stats(struct evlist *evlist);
 
 int perf_stat_process_counter(struct perf_stat_config *config,
 			      struct evsel *counter);
+void perf_stat_merge_counters(struct perf_stat_config *config, struct evlist *evlist);
+
 struct perf_tool;
 union perf_event;
 struct perf_session;
-- 
2.38.0.rc1.362.ged0d419d3c-goog


  parent reply	other threads:[~2022-10-10  5:37 UTC|newest]

Thread overview: 63+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-10  5:35 [RFC/PATCHSET 00/19] perf stat: Cleanup counter aggregation (v1) Namhyung Kim
2022-10-10  5:35 ` [PATCH 01/19] perf tools: Save evsel->pmu in parse_events() Namhyung Kim
2022-10-10 22:21   ` Ian Rogers
2022-10-10  5:35 ` [PATCH 02/19] perf tools: Use pmu info in evsel__is_hybrid() Namhyung Kim
2022-10-10 22:31   ` Ian Rogers
2022-10-11  5:10     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 03/19] perf stat: Use evsel__is_hybrid() more Namhyung Kim
2022-10-10 22:32   ` Ian Rogers
2022-10-10  5:35 ` [PATCH 04/19] perf stat: Add aggr id for global mode Namhyung Kim
2022-10-10 22:46   ` Ian Rogers
2022-10-11 23:08     ` Namhyung Kim
2022-10-12 10:55   ` Jiri Olsa
2022-10-12 16:31     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 05/19] perf stat: Add cpu aggr id for no aggregation mode Namhyung Kim
2022-10-10 22:49   ` Ian Rogers
2022-10-12 10:40   ` Jiri Olsa
2022-10-12 16:27     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 06/19] perf stat: Add 'needs_sort' argument to cpu_aggr_map__new() Namhyung Kim
2022-10-10 22:53   ` Ian Rogers
2022-10-11 23:32     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 07/19] perf stat: Add struct perf_stat_aggr to perf_stat_evsel Namhyung Kim
2022-10-10 23:00   ` Ian Rogers
2022-10-11 23:37     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 08/19] perf stat: Allocate evsel->stats->aggr properly Namhyung Kim
2022-10-10 23:03   ` Ian Rogers
2022-10-11 23:38     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 09/19] perf stat: Aggregate events using evsel->stats->aggr Namhyung Kim
2022-10-10 23:11   ` Ian Rogers
2022-10-11 23:44     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 10/19] perf stat: Aggregate per-thread stats " Namhyung Kim
2022-10-10 23:17   ` Ian Rogers
2022-10-11 23:46     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 11/19] perf stat: Allocate aggr counts for recorded data Namhyung Kim
2022-10-10 23:18   ` Ian Rogers
2022-10-10  5:35 ` [PATCH 12/19] perf stat: Reset aggr counts for each interval Namhyung Kim
2022-10-10 23:20   ` Ian Rogers
2022-10-11 23:48     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 13/19] perf stat: Split process_counters() Namhyung Kim
2022-10-10 23:21   ` Ian Rogers
2022-10-10  5:35 ` Namhyung Kim [this message]
2022-10-10 23:31   ` [PATCH 14/19] perf stat: Add perf_stat_merge_counters() Ian Rogers
2022-10-11 23:55     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 15/19] perf stat: Add perf_stat_process_percore() Namhyung Kim
2022-10-10 23:32   ` Ian Rogers
2022-10-11 23:59     ` Namhyung Kim
2022-10-10  5:35 ` [PATCH 16/19] perf stat: Add perf_stat_process_shadow_stats() Namhyung Kim
2022-10-10 23:36   ` Ian Rogers
2022-10-10  5:35 ` [PATCH 17/19] perf stat: Display event stats using aggr counts Namhyung Kim
2022-10-10 23:38   ` Ian Rogers
2022-10-10  5:35 ` [PATCH 18/19] perf stat: Display percore events properly Namhyung Kim
2022-10-10 23:39   ` Ian Rogers
2022-10-10  5:36 ` [PATCH 19/19] perf stat: Remove unused perf_counts.aggr field Namhyung Kim
2022-10-10 23:40   ` Ian Rogers
2022-10-12  8:41   ` Jiri Olsa
2022-10-12 16:26     ` Namhyung Kim
2022-10-13 20:56       ` [PATCH] perf stat: Init aggr_map when reporting per-process stat Namhyung Kim
2022-10-11  0:25 ` [RFC/PATCHSET 00/19] perf stat: Cleanup counter aggregation (v1) Andi Kleen
2022-10-11  5:38   ` Namhyung Kim
2022-10-11  6:13     ` Ian Rogers
2022-10-12  3:55       ` Namhyung Kim
2022-10-11 11:57     ` Andi Kleen
2022-10-12  3:58       ` Namhyung Kim
  -- strict thread matches above, loose matches on Subject: below --
2022-10-14  6:15 [PATCHSET 00/19] perf stat: Cleanup counter aggregation (v2) Namhyung Kim
2022-10-14  6:15 ` [PATCH 14/19] perf stat: Add perf_stat_merge_counters() Namhyung Kim

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=20221010053600.272854-15-namhyung@kernel.org \
    --to=namhyung@kernel.org \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=ak@linux.intel.com \
    --cc=atrajeev@linux.vnet.ibm.com \
    --cc=irogers@google.com \
    --cc=james.clark@arm.com \
    --cc=jolsa@kernel.org \
    --cc=kan.liang@linux.intel.com \
    --cc=leo.yan@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=zhengjun.xing@linux.intel.com \
    /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).