Messages in this thread Patch in this message | | | From | Ravi Bangoria <> | Subject | [RFC 2/2] perf stat: Fix shadow stats for clock events | Date | Thu, 15 Nov 2018 15:25:33 +0530 |
| |
Commit 0aa802a79469 ("perf stat: Get rid of extra clock display function") introduced scale and unit for clock events. Thus, perf_stat__update_shadow_stats() now saves scaled values of clock events in msecs, instead of original nsecs. But while calculating values of shadow stats we still consider clock event values in nsecs. This results in a wrong shadow stat values. Ex,
# ./perf stat -e task-clock,cycles ls <SNIP> 2.62 msec task-clock:u # 0.624 CPUs utilized 2,501,536 cycles:u # 1250768.000 GHz
Fix this by considering clock events's saved stats in msecs:
# ./perf stat -e task-clock,cycles ls <SNIP> 2.42 msec task-clock:u # 0.754 CPUs utilized 2,338,747 cycles:u # 1.169 GHz
Note: The problem with this approach is, we are losing fractional part while converting nsecs to msecs. This results in a sightly different values of shadow stats.
Reported-by: Anton Blanchard <anton@samba.org> Fixes: 0aa802a79469 ("perf stat: Get rid of extra clock display function") Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> --- tools/perf/util/stat-shadow.c | 13 +++++++------ tools/perf/util/stat.h | 2 +- 2 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c index f0a8cec55c47..5b28b278a24e 100644 --- a/tools/perf/util/stat-shadow.c +++ b/tools/perf/util/stat-shadow.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include <stdio.h> +#include <linux/time64.h> #include "evsel.h" #include "stat.h" #include "color.h" @@ -213,7 +214,7 @@ void perf_stat__update_shadow_stats(struct perf_evsel *counter, u64 count, count *= counter->scale; if (perf_evsel__is_clock(counter)) - update_runtime_stat(st, STAT_NSECS, 0, cpu, count); + update_runtime_stat(st, STAT_MSECS, 0, cpu, count); else if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES)) update_runtime_stat(st, STAT_CYCLES, ctx, cpu, count); else if (perf_stat_evsel__is(counter, CYCLES_IN_TX)) @@ -873,10 +874,10 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, } else if (perf_evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_BACKEND)) { print_stalled_cycles_backend(config, cpu, evsel, avg, out, st); } else if (perf_evsel__match(evsel, HARDWARE, HW_CPU_CYCLES)) { - total = runtime_stat_avg(st, STAT_NSECS, 0, cpu); + total = runtime_stat_avg(st, STAT_MSECS, 0, cpu); if (total) { - ratio = avg / total; + ratio = avg / (total * NSEC_PER_MSEC); print_metric(config, ctxp, NULL, "%8.3f", "GHz", ratio); } else { print_metric(config, ctxp, NULL, NULL, "Ghz", 0); @@ -972,14 +973,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, } else if (evsel->metric_expr) { generic_metric(config, evsel->metric_expr, evsel->metric_events, evsel->name, evsel->metric_name, avg, cpu, out, st); - } else if (runtime_stat_n(st, STAT_NSECS, 0, cpu) != 0) { + } else if (runtime_stat_n(st, STAT_MSECS, 0, cpu) != 0) { char unit = 'M'; char unit_buf[10]; - total = runtime_stat_avg(st, STAT_NSECS, 0, cpu); + total = runtime_stat_avg(st, STAT_MSECS, 0, cpu); if (total) - ratio = 1000.0 * avg / total; + ratio = 1000.0 * avg / (total * NSEC_PER_MSEC); if (ratio < 0.001) { ratio *= 1000; unit = 'K'; diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h index 2f9c9159a364..941ad4b0836c 100644 --- a/tools/perf/util/stat.h +++ b/tools/perf/util/stat.h @@ -62,7 +62,7 @@ enum { enum stat_type { STAT_NONE = 0, - STAT_NSECS, + STAT_MSECS, /* Milliseconds */ STAT_CYCLES, STAT_STALLED_CYCLES_FRONT, STAT_STALLED_CYCLES_BACK, -- 2.17.1
| |