From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id EF646E64AB9 for ; Tue, 3 Dec 2024 15:18:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:List-Subscribe:List-Help :List-Post:List-Archive:List-Unsubscribe:List-Id:Content-Transfer-Encoding: MIME-Version:References:In-Reply-To:Message-ID:Date:Subject:Cc:To:From: Reply-To:Content-Type:Content-ID:Content-Description:Resent-Date:Resent-From: Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=AtfxN7sENpEMTbuvCdRdoHpnn0DF5MS9E6FtTuNVFY4=; b=fGqKpvKlZY+fuX1vDKj5zH0iEJ X9YNsjpx9TFJGJDx44/MnRE353uTEeSc5Bb/tywi4UyAerVoX185RxG7z/6SBSnbcgCKP1E9zPyBg A5sVTXXOqRYmSBsubt0W0IMQlWe2v2rdADhT8GByRipfWxAOFZuyFTgqerYVcoMhDR4UiqDvBF8uC QwD25oU64pwAciodF/E3f4CAS03Jzn6Fc/cpcX03rttqBe2bv85HN5zM7OQ0EWnKvzj1moAmQUFIQ /fPywqk4QHRe22PYj3oJfoJcjoP8ExUeDHA/12dUfV8vDFgu9fx4vyfMY1Jm74BxaGzFQESSutBMB +xUIip7g==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.98 #2 (Red Hat Linux)) id 1tIUfU-00000009u36-365q; Tue, 03 Dec 2024 15:18:28 +0000 Received: from nyc.source.kernel.org ([147.75.193.91]) by bombadil.infradead.org with esmtps (Exim 4.98 #2 (Red Hat Linux)) id 1tIUeU-00000009tt8-1aLt for linux-arm-kernel@lists.infradead.org; Tue, 03 Dec 2024 15:17:28 +0000 Received: from smtp.kernel.org (transwarp.subspace.kernel.org [100.75.92.58]) by nyc.source.kernel.org (Postfix) with ESMTP id 4AE38A416CD; Tue, 3 Dec 2024 15:15:33 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id B0369C4CED8; Tue, 3 Dec 2024 15:17:24 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1733239045; bh=EF4xBJvDmICYpURJnWySEzsRhbId8eDhxP+tp8kHEeg=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=JyHz5r8IEjMtaIrGNgJM3FnF3PnljAxABhWOLs8Ob2qmgMNl4LdqHDGgDyp2tgwkY maEnNSpzsgL6yyEA3STbMraWx/Ax89Dmi72sECfiHOqcBPs3M0MxGMxF6Buu7anHO/ c/TJuD03lJCq9ayY9oYC+4ajn+weCXtHs5617/9g= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, Namhyung Kim , Ian Rogers , Kan Liang , James Clark , Yang Jihong , Dominique Martinet , Colin Ian King , Howard Chu , Ze Gao , Yicong Yang , Weilin Wang , Will Deacon , Mike Leach , Jing Zhang , Yang Li , Leo Yan , ak@linux.intel.com, Athira Rajeev , linux-arm-kernel@lists.infradead.org, Sun Haiyong , John Garry , Sasha Levin Subject: [PATCH 6.11 486/817] perf stat: Uniquify event name improvements Date: Tue, 3 Dec 2024 15:40:58 +0100 Message-ID: <20241203144014.842031727@linuxfoundation.org> X-Mailer: git-send-email 2.47.1 In-Reply-To: <20241203143955.605130076@linuxfoundation.org> References: <20241203143955.605130076@linuxfoundation.org> User-Agent: quilt/0.67 X-stable: review X-Patchwork-Hint: ignore MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20241203_071726_552929_4B46C4C6 X-CRM114-Status: GOOD ( 23.67 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org 6.11-stable review patch. If anyone has any objections, please let me know. ------------------ From: Ian Rogers [ Upstream commit 057f8bfc6f7070577523d1e3081081bbf4229c1c ] Without aggregation on Intel: ``` $ perf stat -e instructions,cycles ... ``` Will use "cycles" for the name of the legacy cycles event but as "instructions" has a sysfs name it will and a "[cpu]" PMU suffix. This often breaks things as the space between the event and the PMU name look like an extra column. The existing uniquify logic was also uniquifying in cases when all events are core and not with uncore events, it was not correctly handling modifiers, etc. Change the logic so that an initial pass that can disable uniquification is run. For individual counters, disable uniquification in more cases such as for consistency with legacy events or for libpfm4 events. Don't use the "[pmu]" style suffix in uniquification, always use "pmu/.../". Change how modifiers/terms are handled in the uniquification so that they look like parse-able events. This fixes "102: perf stat metrics (shadow stat) test:" that has been failing due to "instructions [cpu]" breaking its column/awk logic when values aren't aggregated. This started happening when instructions could match a sysfs rather than a legacy event, so the fixes tag reflects this. Fixes: 617824a7f0f7 ("perf parse-events: Prefer sysfs/JSON hardware events over legacy") Acked-by: Namhyung Kim Signed-off-by: Ian Rogers [ Fix Intel TPEBS counting mode test ] Acked-by: Kan Liang Signed-off-by: James Clark Cc: Yang Jihong Cc: Dominique Martinet Cc: Colin Ian King Cc: Howard Chu Cc: Ze Gao Cc: Yicong Yang Cc: Weilin Wang Cc: Will Deacon Cc: Mike Leach Cc: Jing Zhang Cc: Yang Li Cc: Leo Yan Cc: ak@linux.intel.com Cc: Athira Rajeev Cc: linux-arm-kernel@lists.infradead.org Cc: Sun Haiyong Cc: John Garry Link: https://lore.kernel.org/r/20240926144851.245903-3-james.clark@linaro.org Signed-off-by: Namhyung Kim Signed-off-by: Sasha Levin --- .../perf/tests/shell/test_stat_intel_tpebs.sh | 11 +- tools/perf/util/stat-display.c | 101 ++++++++++++++---- 2 files changed, 85 insertions(+), 27 deletions(-) diff --git a/tools/perf/tests/shell/test_stat_intel_tpebs.sh b/tools/perf/tests/shell/test_stat_intel_tpebs.sh index c60b29add9801..9a11f42d153ca 100755 --- a/tools/perf/tests/shell/test_stat_intel_tpebs.sh +++ b/tools/perf/tests/shell/test_stat_intel_tpebs.sh @@ -8,12 +8,15 @@ grep -q GenuineIntel /proc/cpuinfo || { echo Skipping non-Intel; exit 2; } # Use this event for testing because it should exist in all platforms event=cache-misses:R +# Hybrid platforms output like "cpu_atom/cache-misses/R", rather than as above +alt_name=/cache-misses/R + # Without this cmd option, default value or zero is returned -echo "Testing without --record-tpebs" -result=$(perf stat -e "$event" true 2>&1) -[[ "$result" =~ $event ]] || exit 1 +#echo "Testing without --record-tpebs" +#result=$(perf stat -e "$event" true 2>&1) +#[[ "$result" =~ $event || "$result" =~ $alt_name ]] || exit 1 # In platforms that do not support TPEBS, it should execute without error. echo "Testing with --record-tpebs" result=$(perf stat -e "$event" --record-tpebs -a sleep 0.01 2>&1) -[[ "$result" =~ "perf record" && "$result" =~ $event ]] || exit 1 +[[ "$result" =~ "perf record" && "$result" =~ $event || "$result" =~ $alt_name ]] || exit 1 diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c index ea96e4ebad8c8..cbff43ff8d0fb 100644 --- a/tools/perf/util/stat-display.c +++ b/tools/perf/util/stat-display.c @@ -871,38 +871,66 @@ static void printout(struct perf_stat_config *config, struct outstate *os, static void uniquify_event_name(struct evsel *counter) { - char *new_name; - char *config; - int ret = 0; + const char *name, *pmu_name; + char *new_name, *config; + int ret; - if (counter->uniquified_name || counter->use_config_name || - !counter->pmu_name || !strncmp(evsel__name(counter), counter->pmu_name, - strlen(counter->pmu_name))) + /* The evsel was already uniquified. */ + if (counter->uniquified_name) return; - config = strchr(counter->name, '/'); + /* Avoid checking to uniquify twice. */ + counter->uniquified_name = true; + + /* The evsel has a "name=" config term or is from libpfm. */ + if (counter->use_config_name || counter->is_libpfm_event) + return; + + /* Legacy no PMU event, don't uniquify. */ + if (!counter->pmu || + (counter->pmu->type < PERF_TYPE_MAX && counter->pmu->type != PERF_TYPE_RAW)) + return; + + /* A sysfs or json event replacing a legacy event, don't uniquify. */ + if (counter->pmu->is_core && counter->alternate_hw_config != PERF_COUNT_HW_MAX) + return; + + name = evsel__name(counter); + pmu_name = counter->pmu->name; + /* Already prefixed by the PMU name. */ + if (!strncmp(name, pmu_name, strlen(pmu_name))) + return; + + config = strchr(name, '/'); if (config) { - if (asprintf(&new_name, - "%s%s", counter->pmu_name, config) > 0) { - free(counter->name); - counter->name = new_name; - } - } else { - if (evsel__is_hybrid(counter)) { - ret = asprintf(&new_name, "%s/%s/", - counter->pmu_name, counter->name); + int len = config - name; + + if (config[1] == '/') { + /* case: event// */ + ret = asprintf(&new_name, "%s/%.*s/%s", pmu_name, len, name, config + 2); } else { - ret = asprintf(&new_name, "%s [%s]", - counter->name, counter->pmu_name); + /* case: event/.../ */ + ret = asprintf(&new_name, "%s/%.*s,%s", pmu_name, len, name, config + 1); } + } else { + config = strchr(name, ':'); + if (config) { + /* case: event:.. */ + int len = config - name; - if (ret) { - free(counter->name); - counter->name = new_name; + ret = asprintf(&new_name, "%s/%.*s/%s", pmu_name, len, name, config + 1); + } else { + /* case: event */ + ret = asprintf(&new_name, "%s/%s/", pmu_name, name); } } - - counter->uniquified_name = true; + if (ret > 0) { + free(counter->name); + counter->name = new_name; + } else { + /* ENOMEM from asprintf. */ + counter->uniquified_name = false; + } } static bool hybrid_uniquify(struct evsel *evsel, struct perf_stat_config *config) @@ -1559,6 +1587,31 @@ static void print_cgroup_counter(struct perf_stat_config *config, struct evlist print_metric_end(config, os); } +static void disable_uniquify(struct evlist *evlist) +{ + struct evsel *counter; + struct perf_pmu *last_pmu = NULL; + bool first = true; + + evlist__for_each_entry(evlist, counter) { + /* If PMUs vary then uniquify can be useful. */ + if (!first && counter->pmu != last_pmu) + return; + first = false; + if (counter->pmu) { + /* Allow uniquify for uncore PMUs. */ + if (!counter->pmu->is_core) + return; + /* Keep hybrid event names uniquified for clarity. */ + if (perf_pmus__num_core_pmus() > 1) + return; + } + } + evlist__for_each_entry_continue(evlist, counter) { + counter->uniquified_name = true; + } +} + void evlist__print_counters(struct evlist *evlist, struct perf_stat_config *config, struct target *_target, struct timespec *ts, int argc, const char **argv) @@ -1572,6 +1625,8 @@ void evlist__print_counters(struct evlist *evlist, struct perf_stat_config *conf .first = true, }; + disable_uniquify(evlist); + if (config->iostat_run) evlist->selected = evlist__first(evlist); -- 2.43.0