[PATCH] tools/perf: Fix aggr_printout to display cpu field irrespective of core value

Arnaldo Carvalho de Melo acme at kernel.org
Wed Oct 5 23:28:52 AEDT 2022


Em Wed, Oct 05, 2022 at 10:23:39AM +0530, Athira Rajeev escreveu:
> 
> 
> > On 04-Oct-2022, at 11:44 PM, Arnaldo Carvalho de Melo <acme at kernel.org> wrote:
> > 
> > Em Tue, Oct 04, 2022 at 03:14:27PM -0300, Arnaldo Carvalho de Melo escreveu:
> >> Em Tue, Oct 04, 2022 at 07:49:21AM -0700, Ian Rogers escreveu:
> >>> On Tue, Oct 4, 2022, 12:06 AM Athira Rajeev <atrajeev at linux.vnet.ibm.com>
> >>>> Thanks for helping with testing. Can I add your Tested-by for the patch ?
> >> 
> >>> Yep.
> >> 
> >>> Tested-by: Ian Rogers <irogers at google.com>
> > 
> > 
> > Thanks, applied.
> > 
> > - Arnaldo
> 
> Hi Arnaldo,
> 
> Looks like you have taken change to remove id.core check:
> https://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git/commit/?h=tmp.perf/core&id=db83f447b323958cdc5fedcf2134effb2ec9a6fe
> 
> But the patch that has to go in is :
> "[PATCH] tools/perf: Fix cpu check to use id.cpu.cpu in ggr_printout"
> which is tested by Ian and "pasted" by me in same mail thread.
> 
> Re-pasting here for reference:
> 
> >From 4dd98d953940deb2f85176cb6b4ecbfd18dbdbf9 Mon Sep 17 00:00:00 2001
> From: Athira Rajeev <atrajeev at linux.vnet.ibm.com>
> Date: Mon, 3 Oct 2022 15:47:27 +0530
> Subject: [PATCH] tools/perf: Fix cpu check to use id.cpu.cpu in aggr_printout
> 
> perf stat has options to aggregate the counts in different
> modes like per socket, per core etc. The function "aggr_printout"
> in util/stat-display.c which is used to print the aggregates,
> has a check for cpu in case of AGGR_NONE. This check was
> originally using condition : "if (id.cpu.cpu > -1)". But
> this got changed after commit df936cadfb58 ("perf stat: Add
> JSON output option"), which added option to output json format
> for different aggregation modes. After this commit, the
> check in "aggr_printout" is using "if (id.core > -1)".
> 
> The old code was using "id.cpu.cpu > -1" while the new code
> is using "id.core > -1". But since the value printed is
> id.cpu.cpu, fix this check to use cpu and not core.
> 
> Signed-off-by: Athira Rajeev <atrajeev at linux.vnet.ibm.com>
> Suggested-by: James Clark <james.clark at arm.com>
> Suggested-by: Ian Rogers <irogers at google.com>
> ---
> tools/perf/util/stat-display.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c
> index b82844cb0ce7..cf28020798ec 100644
> --- a/tools/perf/util/stat-display.c
> +++ b/tools/perf/util/stat-display.c
> @@ -168,7 +168,7 @@ static void aggr_printout(struct perf_stat_config *config,
> 					id.socket,
> 					id.die,
> 					id.core);
> -			} else if (id.core > -1) {
> +			} else if (id.cpu.cpu > -1) {
> 				fprintf(config->output, "\"cpu\" : \"%d\", ",
> 					id.cpu.cpu);
> 			}
> @@ -179,7 +179,7 @@ static void aggr_printout(struct perf_stat_config *config,
> 					id.die,
> 					config->csv_output ? 0 : -3,
> 					id.core, config->csv_sep);
> -			} else if (id.core > -1) {
> +			} else if (id.cpu.cpu > -1) {
> 				fprintf(config->output, "CPU%*d%s",
> 					config->csv_output ? 0 : -7,
> 					id.cpu.cpu, config->csv_sep);
> -- 
> 2.31.1
> 
> If it is confusing, shall I send it as a separate patch along with Tested-by from Ian ?

I'll have to do this by hand, tried pointing b4 to this message and it
picked the old one, also tried to save the message and apply by hand,
its mangled.

- Arnaldo
> 
> Thanks
> Athira
> 
> > 

-- 

- Arnaldo


More information about the Linuxppc-dev mailing list