| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <20211120030053.2649499-1-irogers@google.com>
Date: Fri, 19 Nov 2021 19:00:53 -0800
From: Ian Rogers <irogers@...gle.com>
To: Andi Kleen <ak@...ux.intel.com>, Jiri Olsa <jolsa@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
John Garry <john.garry@...wei.com>,
Kajol Jain <kjain@...ux.ibm.com>,
"Paul A . Clarke" <pc@...ibm.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
Kan Liang <kan.liang@...ux.intel.com>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org
Cc: eranian@...gle.com, Ian Rogers <irogers@...gle.com>
Subject: [PATCH] perf metric: Reduce multiplexing with duration_time
It is common to use the same counters with and without
duration_time. The ID sharing code treats duration_time as if it
were a hardware event placed in the same group. This causes
unnecessary multiplexing such as in the following example where
l3_cache_access isn't shared:
$ perf stat -M l3 -a sleep 1
Performance counter stats for 'system wide':
3,117,007 l3_cache_miss # 199.5 MB/s l3_rd_bw
# 43.6 % l3_hits
# 56.4 % l3_miss (50.00%)
5,526,447 l3_cache_access (50.00%)
5,392,435 l3_cache_access # 5389191.2 access/s l3_access_rate (50.00%)
1,000,601,901 ns duration_time
1.000601901 seconds time elapsed
Fix this by placing duration_time in all groups unless metric
sharing has been disabled on the command line:
$ perf stat -M l3 -a sleep 1
Performance counter stats for 'system wide':
3,597,972 l3_cache_miss # 230.3 MB/s l3_rd_bw
# 48.0 % l3_hits
# 52.0 % l3_miss
6,914,459 l3_cache_access # 6909935.9 access/s l3_access_rate
1,000,654,579 ns duration_time
1.000654579 seconds time elapsed
$ perf stat --metric-no-merge -M l3 -a sleep 1
Performance counter stats for 'system wide':
3,501,834 l3_cache_miss # 53.5 % l3_miss (24.99%)
6,548,173 l3_cache_access (24.99%)
3,417,622 l3_cache_miss # 45.7 % l3_hits (25.04%)
6,294,062 l3_cache_access (25.04%)
5,923,238 l3_cache_access # 5919688.1 access/s l3_access_rate (24.99%)
1,000,599,683 ns duration_time
3,607,486 l3_cache_miss # 230.9 MB/s l3_rd_bw (49.97%)
1.000599683 seconds time elapsed
Signed-off-by: Ian Rogers <irogers@...gle.com>
---
tools/perf/util/metricgroup.c | 26 ++++++++++++++++++--------
1 file changed, 18 insertions(+), 8 deletions(-)
diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c
index fffe02aae3ed..145c59ba2ddc 100644
--- a/tools/perf/util/metricgroup.c
+++ b/tools/perf/util/metricgroup.c
@@ -1299,14 +1299,16 @@ static int build_combined_expr_ctx(const struct list_head *metric_list,
/**
* parse_ids - Build the event string for the ids and parse them creating an
* evlist. The encoded metric_ids are decoded.
+ * @metric_no_merge: is metric sharing explicitly disabled.
* @fake_pmu: used when testing metrics not supported by the current CPU.
* @ids: the event identifiers parsed from a metric.
* @modifier: any modifiers added to the events.
* @has_constraint: false if events should be placed in a weak group.
* @out_evlist: the created list of events.
*/
-static int parse_ids(struct perf_pmu *fake_pmu, struct expr_parse_ctx *ids,
- const char *modifier, bool has_constraint, struct evlist **out_evlist)
+static int parse_ids(bool metric_no_merge, struct perf_pmu *fake_pmu,
+ struct expr_parse_ctx *ids, const char *modifier,
+ bool has_constraint, struct evlist **out_evlist)
{
struct parse_events_error parse_error;
struct evlist *parsed_evlist;
@@ -1314,12 +1316,19 @@ static int parse_ids(struct perf_pmu *fake_pmu, struct expr_parse_ctx *ids,
int ret;
*out_evlist = NULL;
- if (hashmap__size(ids->ids) == 0) {
+ if (!metric_no_merge || hashmap__size(ids->ids) == 0) {
char *tmp;
/*
- * No ids/events in the expression parsing context. Events may
- * have been removed because of constant evaluation, e.g.:
- * event1 if #smt_on else 0
+ * We may fail to share events between metrics because
+ * duration_time isn't present in one metric. For example, a
+ * ratio of cache misses doesn't need duration_time but the same
+ * events may be used for a misses per second. Events without
+ * sharing implies multiplexing, that is best avoided, so place
+ * duration_time in every group.
+ *
+ * Also, there may be no ids/events in the expression parsing
+ * context because of constant evaluation, e.g.:
+ * event1 if #smt_on else 0
* Add a duration_time event to avoid a parse error on an empty
* string.
*/
@@ -1387,7 +1396,8 @@ static int parse_groups(struct evlist *perf_evlist, const char *str,
ret = build_combined_expr_ctx(&metric_list, &combined);
if (!ret && combined && hashmap__size(combined->ids)) {
- ret = parse_ids(fake_pmu, combined, /*modifier=*/NULL,
+ ret = parse_ids(metric_no_merge, fake_pmu, combined,
+ /*modifier=*/NULL,
/*has_constraint=*/true,
&combined_evlist);
}
@@ -1435,7 +1445,7 @@ static int parse_groups(struct evlist *perf_evlist, const char *str,
}
}
if (!metric_evlist) {
- ret = parse_ids(fake_pmu, m->pctx, m->modifier,
+ ret = parse_ids(metric_no_merge, fake_pmu, m->pctx, m->modifier,
m->has_constraint, &m->evlist);
if (ret)
goto out;
--
2.34.0.rc2.393.gf8c9666880-goog
Powered by blists - more mailing lists