[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240324224720.1345309-459-sashal@kernel.org>
Date: Sun, 24 Mar 2024 18:43:04 -0400
From: Sasha Levin <sashal@...nel.org>
To: linux-kernel@...r.kernel.org,
stable@...r.kernel.org
Cc: Ian Rogers <irogers@...gle.com>,
Kan Liang <kan.liang@...ux.intel.com>,
K Prateek Nayak <kprateek.nayak@....com>,
James Clark <james.clark@....com>,
Kaige Ye <ye@...ge.org>,
John Garry <john.g.garry@...cle.com>,
Namhyung Kim <namhyung@...nel.org>,
Sasha Levin <sashal@...nel.org>
Subject: [PATCH 6.7 458/713] perf metric: Don't remove scale from counts
From: Ian Rogers <irogers@...gle.com>
[ Upstream commit 6d6be5eb45b423a37d746d3ee0fd0c78f76ead9f ]
Counts were switched from the scaled saved value form to the
aggregated count to avoid double accounting. When this happened the
removing of scaling for a count should have been removed, however, it
wasn't and this wasn't observed as it normally doesn't matter because
a counter's scale is 1. A problem was observed with RAPL events that
are scaled.
Fixes: 37cc8ad77cf8 ("perf metric: Directly use counts rather than saved_value")
Signed-off-by: Ian Rogers <irogers@...gle.com>
Reviewed-by: Kan Liang <kan.liang@...ux.intel.com>
Cc: K Prateek Nayak <kprateek.nayak@....com>
Cc: James Clark <james.clark@....com>
Cc: Kaige Ye <ye@...ge.org>
Cc: John Garry <john.g.garry@...cle.com>
Signed-off-by: Namhyung Kim <namhyung@...nel.org>
Link: https://lore.kernel.org/r/20240209204947.3873294-5-irogers@google.com
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
tools/perf/util/stat-shadow.c | 7 +------
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index e31426167852a..cf573ff3fa84f 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -414,12 +414,7 @@ static int prepare_metric(struct evsel **metric_events,
val = NAN;
source_count = 0;
} else {
- /*
- * If an event was scaled during stat gathering,
- * reverse the scale before computing the
- * metric.
- */
- val = aggr->counts.val * (1.0 / metric_events[i]->scale);
+ val = aggr->counts.val;
source_count = evsel__source_count(metric_events[i]);
}
}
--
2.43.0
Powered by blists - more mailing lists