From: Jin Yao <yao.jin@linux.intel.com>
To: acme@kernel.org, jolsa@kernel.org, peterz@infradead.org,
mingo@redhat.com, alexander.shishkin@linux.intel.com
Cc: Linux-kernel@vger.kernel.org, ak@linux.intel.com,
kan.liang@intel.com, yao.jin@intel.com,
Jin Yao <yao.jin@linux.intel.com>
Subject: [PATCH v5 14/26] perf stat: Add default hybrid events
Date: Fri, 23 Apr 2021 13:35:29 +0800 [thread overview]
Message-ID: <20210423053541.12521-15-yao.jin@linux.intel.com> (raw)
In-Reply-To: <20210423053541.12521-1-yao.jin@linux.intel.com>
Previously if '-e' is not specified in perf stat, some software events
and hardware events are added to evlist by default.
Before:
# perf stat -a -- sleep 1
Performance counter stats for 'system wide':
24,044.40 msec cpu-clock # 23.946 CPUs utilized
99 context-switches # 4.117 /sec
24 cpu-migrations # 0.998 /sec
3 page-faults # 0.125 /sec
7,000,244 cycles # 0.000 GHz
2,955,024 instructions # 0.42 insn per cycle
608,941 branches # 25.326 K/sec
31,991 branch-misses # 5.25% of all branches
1.004106859 seconds time elapsed
Among the events, cycles, instructions, branches and branch-misses
are hardware events.
One hybrid platform, two hardware events are created for one
hardware event.
cpu_core/cycles/,
cpu_atom/cycles/,
cpu_core/instructions/,
cpu_atom/instructions/,
cpu_core/branches/,
cpu_atom/branches/,
cpu_core/branch-misses/,
cpu_atom/branch-misses/
These events would be added to evlist on hybrid platform.
Since parse_events() has been supported to create two hardware events
for one event on hybrid platform, so we just use parse_events(evlist,
"cycles,instructions,branches,branch-misses") to create the default
events and add them to evlist.
After:
# perf stat -a -- sleep 1
Performance counter stats for 'system wide':
24,043.99 msec cpu-clock # 23.991 CPUs utilized
139 context-switches # 5.781 /sec
25 cpu-migrations # 1.040 /sec
6 page-faults # 0.250 /sec
10,381,751 cpu_core/cycles/ # 431.782 K/sec
1,264,216 cpu_atom/cycles/ # 52.579 K/sec
3,406,958 cpu_core/instructions/ # 141.697 K/sec
414,588 cpu_atom/instructions/ # 17.243 K/sec
705,149 cpu_core/branches/ # 29.327 K/sec
82,358 cpu_atom/branches/ # 3.425 K/sec
40,821 cpu_core/branch-misses/ # 1.698 K/sec
9,086 cpu_atom/branch-misses/ # 377.891 /sec
1.002228863 seconds time elapsed
We can see two events are created for one hardware event.
One TODO is, the shadow stats looks a bit different, now it's just
'M/sec'.
The perf_stat__update_shadow_stats and perf_stat__print_shadow_stats
need to be improved in future if we want to get the original shadow
stats.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
---
v5:
- Do the same way like when topdown calls parse events for checking
result and displayt the error.
v4:
- No change.
tools/perf/builtin-stat.c | 28 ++++++++++++++++++++++++++++
1 file changed, 28 insertions(+)
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 1255af4751c2..3ab4069ff8f0 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -1626,6 +1626,12 @@ static int add_default_attributes(void)
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES },
+};
+ struct perf_event_attr default_sw_attrs[] = {
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS },
};
/*
@@ -1863,6 +1869,28 @@ static int add_default_attributes(void)
}
if (!evsel_list->core.nr_entries) {
+ if (perf_pmu__has_hybrid()) {
+ const char *hybrid_str = "cycles,instructions,branches,branch-misses";
+
+ if (target__has_cpu(&target))
+ default_sw_attrs[0].config = PERF_COUNT_SW_CPU_CLOCK;
+
+ if (evlist__add_default_attrs(evsel_list,
+ default_sw_attrs) < 0) {
+ return -1;
+ }
+
+ err = parse_events(evsel_list, hybrid_str, &errinfo);
+ if (err) {
+ fprintf(stderr,
+ "Cannot set up hybrid events %s: %d\n",
+ hybrid_str, err);
+ parse_events_print_error(&errinfo, hybrid_str);
+ return -1;
+ }
+ return err;
+ }
+
if (target__has_cpu(&target))
default_attrs0[0].config = PERF_COUNT_SW_CPU_CLOCK;
--
2.17.1
next prev parent reply other threads:[~2021-04-23 5:37 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-23 5:35 [PATCH v5 00/26] perf tool: AlderLake hybrid support series 1 Jin Yao
2021-04-23 5:35 ` [PATCH v5 01/26] tools headers uapi: Update tools's copy of linux/perf_event.h Jin Yao
2021-04-23 5:35 ` [PATCH v5 02/26] perf jevents: Support unit value "cpu_core" and "cpu_atom" Jin Yao
2021-04-23 5:35 ` [PATCH v5 03/26] perf pmu: Simplify arguments of __perf_pmu__new_alias Jin Yao
2021-04-23 5:35 ` [PATCH v5 04/26] perf pmu: Save pmu name Jin Yao
2021-04-23 5:35 ` [PATCH v5 05/26] perf pmu: Save detected hybrid pmus to a global pmu list Jin Yao
2021-04-23 5:35 ` [PATCH v5 06/26] perf pmu: Add hybrid helper functions Jin Yao
2021-04-23 5:35 ` [PATCH v5 07/26] perf stat: Uniquify hybrid event name Jin Yao
2021-04-23 5:35 ` [PATCH v5 08/26] perf parse-events: Create two hybrid hardware events Jin Yao
2021-04-23 5:35 ` [PATCH v5 09/26] perf parse-events: Create two hybrid cache events Jin Yao
2021-04-23 5:35 ` [PATCH v5 10/26] perf parse-events: Create two hybrid raw events Jin Yao
2021-04-23 5:35 ` [PATCH v5 11/26] perf parse-events: Compare with hybrid pmu name Jin Yao
2021-04-23 5:35 ` [PATCH v5 12/26] perf parse-events: Support event inside hybrid pmu Jin Yao
2021-04-25 16:17 ` Jiri Olsa
2021-04-26 0:56 ` Jin, Yao
2021-04-26 21:34 ` Jiri Olsa
2021-04-23 5:35 ` [PATCH v5 13/26] perf record: Create two hybrid 'cycles' events by default Jin Yao
2021-04-23 5:35 ` Jin Yao [this message]
2021-04-23 5:35 ` [PATCH v5 15/26] perf stat: Filter out unmatched aggregation for hybrid event Jin Yao
2021-04-23 5:35 ` [PATCH v5 16/26] perf stat: Warn group events from different hybrid PMU Jin Yao
2021-04-23 5:35 ` [PATCH v5 17/26] perf record: Uniquify hybrid event name Jin Yao
2021-04-23 5:35 ` [PATCH v5 18/26] perf tests: Add hybrid cases for 'Parse event definition strings' test Jin Yao
2021-04-23 5:35 ` [PATCH v5 19/26] perf tests: Add hybrid cases for 'Roundtrip evsel->name' test Jin Yao
2021-04-23 5:35 ` [PATCH v5 20/26] perf tests: Skip 'Setup struct perf_event_attr' test for hybrid Jin Yao
2021-04-23 5:35 ` [PATCH v5 21/26] perf tests: Support 'Track with sched_switch' " Jin Yao
2021-04-23 5:35 ` [PATCH v5 22/26] perf tests: Support 'Parse and process metrics' " Jin Yao
2021-04-23 5:35 ` [PATCH v5 23/26] perf tests: Support 'Session topology' " Jin Yao
2021-04-23 5:35 ` [PATCH v5 24/26] perf tests: Support 'Convert perf time to TSC' " Jin Yao
2021-04-23 5:35 ` [PATCH v5 25/26] perf tests: Skip 'perf stat metrics (shadow stat) test' " Jin Yao
2021-04-23 5:35 ` [PATCH v5 26/26] perf Documentation: Document intel-hybrid support Jin Yao
2021-04-26 20:41 ` [PATCH v5 00/26] perf tool: AlderLake hybrid support series 1 Arnaldo Carvalho de Melo
2021-04-27 0:48 ` Jin, Yao
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=20210423053541.12521-15-yao.jin@linux.intel.com \
--to=yao.jin@linux.intel.com \
--cc=Linux-kernel@vger.kernel.org \
--cc=acme@kernel.org \
--cc=ak@linux.intel.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=jolsa@kernel.org \
--cc=kan.liang@intel.com \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=yao.jin@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).