Commit 82bf311e authored by Jiri Olsa's avatar Jiri Olsa Committed by Arnaldo Carvalho de Melo
Browse files

perf stat: Use group read for event groups 



Make perf stat use  group read if there  are groups defined. The group
read will get the values for all member of groups within a single
syscall instead of calling read syscall for every event.

We can see considerable less amount of kernel cycles spent on single
group read, than reading each event separately, like for following perf
stat command:

  # perf stat -e {cycles,instructions} -I 10 -a sleep 1

Monitored with "perf stat -r 5 -e '{cycles:u,cycles:k}'"

Before:

        24,325,676      cycles:u
       297,040,775      cycles:k

       1.038554134 seconds time elapsed

After:
        25,034,418      cycles:u
       158,256,395      cycles:k

       1.036864497 seconds time elapsed

The perf_evsel__open fallback changes contributed by Andi Kleen.

Signed-off-by: default avatarJiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20170726120206.9099-4-jolsa@kernel.org


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
parent f7794d52

tools/perf/builtin-stat.c

+27 −3
Original line number Diff line number Diff line
@@ -213,10 +213,20 @@ static void perf_stat__reset_stats(void) 
static int create_perf_stat_counter(struct perf_evsel *evsel)

{

	struct perf_event_attr *attr = &evsel->attr;

	struct perf_evsel *leader = evsel->leader;



	if (stat_config.scale)

	if (stat_config.scale) {

		attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED |

				    PERF_FORMAT_TOTAL_TIME_RUNNING;

	}



	/*

	 * The event is part of non trivial group, let's enable

	 * the group read (for leader) and ID retrieval for all

	 * members.

	 */

	if (leader->nr_members > 1)

		attr->read_format |= PERF_FORMAT_ID|PERF_FORMAT_GROUP;



	attr->inherit = !no_inherit;
@@ -333,13 +343,21 @@ static int read_counter(struct perf_evsel *counter) 
			struct perf_counts_values *count;



			count = perf_counts(counter->counts, cpu, thread);

			if (perf_evsel__read(counter, cpu, thread, count)) {



			/*

			 * The leader's group read loads data into its group members

			 * (via perf_evsel__read_counter) and sets threir count->loaded.

			 */

			if (!count->loaded &&

			    perf_evsel__read_counter(counter, cpu, thread)) {

				counter->counts->scaled = -1;

				perf_counts(counter->counts, cpu, thread)->ena = 0;

				perf_counts(counter->counts, cpu, thread)->run = 0;

				return -1;

			}



			count->loaded = false;



			if (STAT_RECORD) {

				if (perf_evsel__write_stat_event(counter, cpu, thread, count)) {

					pr_err("failed to write stat event\n");
@@ -559,6 +577,11 @@ static int store_counter_ids(struct perf_evsel *counter) 
	return __store_counter_ids(counter, cpus, threads);

}



static bool perf_evsel__should_store_id(struct perf_evsel *counter)

{

	return STAT_RECORD || counter->attr.read_format & PERF_FORMAT_ID;

}



static int __run_perf_stat(int argc, const char **argv)

{

	int interval = stat_config.interval;
@@ -631,7 +654,8 @@ try_again: 
		if (l > unit_width)

			unit_width = l;



		if (STAT_RECORD && store_counter_ids(counter))

		if (perf_evsel__should_store_id(counter) &&

		    store_counter_ids(counter))

			return -1;

	}

tools/perf/util/counts.h

+1 −0
Original line number Diff line number Diff line
@@ -12,6 +12,7 @@ struct perf_counts_values { 
		};

		u64 values[3];

	};

	bool	loaded;

};



struct perf_counts {

tools/perf/util/evsel.c

+10 −0
Original line number Diff line number Diff line
@@ -49,6 +49,7 @@ static struct { 
	bool clockid_wrong;

	bool lbr_flags;

	bool write_backward;

	bool group_read;

} perf_missing_features;



static clockid_t clockid;
@@ -1321,6 +1322,7 @@ perf_evsel__set_count(struct perf_evsel *counter, int cpu, int thread, 
	count->val    = val;

	count->ena    = ena;

	count->run    = run;

	count->loaded = true;

}



static int
@@ -1677,6 +1679,8 @@ fallback_missing_features: 
	if (perf_missing_features.lbr_flags)

		evsel->attr.branch_sample_type &= ~(PERF_SAMPLE_BRANCH_NO_FLAGS |

				     PERF_SAMPLE_BRANCH_NO_CYCLES);

	if (perf_missing_features.group_read && evsel->attr.inherit)

		evsel->attr.read_format &= ~(PERF_FORMAT_GROUP|PERF_FORMAT_ID);

retry_sample_id:

	if (perf_missing_features.sample_id_all)

		evsel->attr.sample_id_all = 0;
@@ -1832,6 +1836,12 @@ try_fallback: 
		perf_missing_features.lbr_flags = true;

		pr_debug2("switching off branch sample type no (cycles/flags)\n");

		goto fallback_missing_features;

	} else if (!perf_missing_features.group_read &&

		    evsel->attr.inherit &&

		   (evsel->attr.read_format & PERF_FORMAT_GROUP)) {

		perf_missing_features.group_read = true;

		pr_debug2("switching off group read\n");

		goto fallback_missing_features;

	}

out_close:

	do {

CRA Git | Maintained and supported by SUSTech CRA and CCSE