From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751805AbeEBPU3 (ORCPT ); Wed, 2 May 2018 11:20:29 -0400 Received: from mail.kernel.org ([198.145.29.99]:46162 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751687AbeEBPUW (ORCPT ); Wed, 2 May 2018 11:20:22 -0400 From: Arnaldo Carvalho de Melo To: Ingo Molnar Cc: Clark Williams , linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org, Jiri Olsa , Alexander Shishkin , Andi Kleen , David Ahern , Namhyung Kim , Peter Zijlstra , Arnaldo Carvalho de Melo Subject: [PATCH 05/52] perf stat: Display time in precision based on std deviation Date: Wed, 2 May 2018 12:18:55 -0300 Message-Id: <20180502151942.20542-6-acme@kernel.org> X-Mailer: git-send-email 2.14.3 In-Reply-To: <20180502151942.20542-1-acme@kernel.org> References: <20180502151942.20542-1-acme@kernel.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Jiri Olsa Ingo suggested to display elapsed time for multirun workload (perf stat -e) with precision based on the precision of the standard deviation. In his own words: > This output is a slightly bit misleading: > Performance counter stats for 'make -j128' (10 runs): > 27.988995256 seconds time elapsed ( +- 0.39% ) > The 9 significant digits in the result, while only 1 is valid, suggests accuracy > where none exists. > It would be better if 'perf stat' would display elapsed time with a precision > adjusted to stddev, it should display at most 2 more significant digits than > the stddev inaccuracy. > I.e. in the above case 0.39% is 0.109, so we only have accuracy for 1 digit, and > so we should only display 3: > 27.988 seconds time elapsed ( +- 0.39% ) Plus a suggestion about the output, which is small enough and connected with the above change that I merged both changes together. > Small output style nit - I think it would be nice if with --repeat the stddev was > also displayed in absolute values, besides percentage: > > 27.988 +- 0.109 seconds time elapsed ( +- 0.39% ) The output is now: Performance counter stats for './perf bench sched pipe' (5 runs): SNIP 13.3667 +- 0.0256 seconds time elapsed ( +- 0.19% ) Suggested-by: Ingo Molnar Signed-off-by: Jiri Olsa Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20180423090823.32309-7-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-stat.c | 30 ++++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index f17dc601b0f3..3a94647ef6bc 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -1764,19 +1764,37 @@ static void print_header(int argc, const char **argv) } } +static int get_precision(double num) +{ + if (num > 1) + return 0; + + return lround(ceil(-log10(num))); +} + static void print_footer(void) { + double avg = avg_stats(&walltime_nsecs_stats) / NSEC_PER_SEC; FILE *output = stat_config.output; int n; if (!null_run) fprintf(output, "\n"); - fprintf(output, " %17.9f seconds time elapsed", - avg_stats(&walltime_nsecs_stats) / NSEC_PER_SEC); - if (run_count > 1) { - fprintf(output, " "); - print_noise_pct(stddev_stats(&walltime_nsecs_stats), - avg_stats(&walltime_nsecs_stats)); + + if (run_count == 1) { + fprintf(output, " %17.9f seconds time elapsed", avg); + } else { + double sd = stddev_stats(&walltime_nsecs_stats) / NSEC_PER_SEC; + /* + * Display at most 2 more significant + * digits than the stddev inaccuracy. + */ + int precision = get_precision(sd) + 2; + + fprintf(output, " %17.*f +- %.*f seconds time elapsed", + precision, avg, precision, sd); + + print_noise_pct(sd, avg); } fprintf(output, "\n\n"); -- 2.14.3