From nobody Sat May 18 18:58:07 2024 Received: from szxga05-in.huawei.com (szxga05-in.huawei.com [45.249.212.191]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 74F621D529; Thu, 8 Feb 2024 02:44:50 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=45.249.212.191 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1707360294; cv=none; b=U6ZhtgneyFEtBfgZHvyWJSyCIn2Lh8UeWjrs2wWHugaSXU0M6s9CFZUSjJl7EKbByrU7Vc6N5m2V+9114hRjK9y5Yt+XJTcjmi7bSNZM+pJQUb339f4+pBz6H5AB8Qc/XCyNpQMx7f+SjvUbkkhSzGnGs56yDqqMQnz0TQQAttA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1707360294; c=relaxed/simple; bh=M9fgCDd6U2+FRcl6NwuV2JdY4MZ6yRqfqpYTR/O0k6E=; h=From:To:CC:Subject:Date:Message-ID:MIME-Version:Content-Type; b=ABfao8BLMluhV+TMOun24N7IaIx3ejykQY5PzaqnUTQ7IIcwVrNjfI47e5cIFrnAebZ1y6i4l3AOsAj6JjoVe/VzzsxK8nDz/aJOWm8FxkNTqQBSJVBbM/oVXR2sxJY+kEnucG7H3jeV8tbgrL/3uVxf8/t8Rw7264jDn8w/mrg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=huawei.com; spf=pass smtp.mailfrom=huawei.com; arc=none smtp.client-ip=45.249.212.191 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=huawei.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=huawei.com Received: from mail.maildlp.com (unknown [172.19.88.234]) by szxga05-in.huawei.com (SkyGuard) with ESMTP id 4TVh9Q4gdbz1gyjX; Thu, 8 Feb 2024 10:42:50 +0800 (CST) Received: from canpemm500009.china.huawei.com (unknown [7.192.105.203]) by mail.maildlp.com (Postfix) with ESMTPS id A1C1F140444; Thu, 8 Feb 2024 10:44:47 +0800 (CST) Received: from localhost.localdomain (10.50.165.33) by canpemm500009.china.huawei.com (7.192.105.203) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.35; Thu, 8 Feb 2024 10:44:47 +0800 From: Yicong Yang To: , , , , , , , , , , , CC: , , <21cnbao@gmail.com>, , , , , , , Tim Chen Subject: [PATCH v5] perf stat: Support per-cluster aggregation Date: Thu, 8 Feb 2024 10:40:26 +0800 Message-ID: <20240208024026.2691-1-yangyicong@huawei.com> X-Mailer: git-send-email 2.31.0 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: dggems701-chm.china.huawei.com (10.3.19.178) To canpemm500009.china.huawei.com (7.192.105.203) Content-Type: text/plain; charset="utf-8" From: Yicong Yang Some platforms have 'cluster' topology and CPUs in the cluster will share resources like L3 Cache Tag (for HiSilicon Kunpeng SoC) or L2 cache (for Intel Jacobsville). Currently parsing and building cluster topology have been supported since [1]. perf stat has already supported aggregation for other topologies like die or socket, etc. It'll be useful to aggregate per-cluster to find problems like L3T bandwidth contention. This patch add support for "--per-cluster" option for per-cluster aggregation. Also update the docs and related test. The output will be like: [root@localhost tmp]# perf stat -a -e LLC-load --per-cluster -- sleep 5 Performance counter stats for 'system wide': S56-D0-CLS158 4 1,321,521,570 LLC-load S56-D0-CLS594 4 794,211,453 LLC-load S56-D0-CLS1030 4 41,623 LLC-load S56-D0-CLS1466 4 41,646 LLC-load S56-D0-CLS1902 4 16,863 LLC-load S56-D0-CLS2338 4 15,721 LLC-load S56-D0-CLS2774 4 22,671 LLC-load [...] On a legacy system without cluster or cluster support, the output will be look like: [root@localhost perf]# perf stat -a -e cycles --per-cluster -- sleep 1 Performance counter stats for 'system wide': S56-D0-CLS0 64 18,011,485 cycles S7182-D0-CLS0 64 16,548,835 cycles Note that this patch doesn't mix the cluster information in the outputs of --per-core to avoid breaking any tools/scripts using it. Note that perf recently supports "--per-cache" aggregation, but it's not the same with the cluster although cluster CPUs may share some cache resources. For example on my machine all clusters within a die share the same L3 cache: $ cat /sys/devices/system/cpu/cpu0/cache/index3/shared_cpu_list 0-31 $ cat /sys/devices/system/cpu/cpu0/topology/cluster_cpus_list 0-3 [1] commit c5e22feffdd7 ("topology: Represent clusters of CPUs within a die= ") Tested-by: Jie Zhan Reviewed-by: Tim Chen Reviewed-by: Ian Rogers Signed-off-by: Yicong Yang --- Change since v4: - Add the Tag from Ian, Thanks. - Fix the code style in cpu__get_cluster_id(), per Ian Link: https://lore.kernel.org/linux-perf-users/20240206082016.22292-1-yangy= icong@huawei.com/ Change since v3: - Rebase on v6.7-rc4 and resolve the conflicts Link: https://lore.kernel.org/all/20230404104951.27537-1-yangyicong@huawei.= com/ Change since v2: - Use 0 as cluster ID on legacy system without cluster support, keep consis= tenct with what --per-die does. Link: https://lore.kernel.org/all/20230328112717.19573-1-yangyicong@huawei.= com/ Change since v1: - Provides the information about how to map the cluster to the CPUs in the = manual - Thanks the review from Tim and test from Jie. Link: https://lore.kernel.org/all/20230313085911.61359-1-yangyicong@huawei.= com/ tools/perf/Documentation/perf-stat.txt | 11 ++++ tools/perf/builtin-stat.c | 52 +++++++++++++++++-- .../tests/shell/lib/perf_json_output_lint.py | 4 +- tools/perf/tests/shell/lib/stat_output.sh | 12 +++++ tools/perf/tests/shell/stat+csv_output.sh | 2 + tools/perf/tests/shell/stat+json_output.sh | 13 +++++ tools/perf/tests/shell/stat+std_output.sh | 2 + tools/perf/util/cpumap.c | 33 +++++++++++- tools/perf/util/cpumap.h | 19 +++++-- tools/perf/util/env.h | 1 + tools/perf/util/stat-display.c | 13 +++++ tools/perf/util/stat.h | 1 + 12 files changed, 154 insertions(+), 9 deletions(-) diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentat= ion/perf-stat.txt index 5af2e432b54f..29756a87ab6f 100644 --- a/tools/perf/Documentation/perf-stat.txt +++ b/tools/perf/Documentation/perf-stat.txt @@ -308,6 +308,14 @@ use --per-die in addition to -a. (system-wide). The o= utput includes the die number and the number of online processors on that die. This is useful to gauge the amount of aggregation. =20 +--per-cluster:: +Aggregate counts per processor cluster for system-wide mode measurement. = This +is a useful mode to detect imbalance between clusters. To enable this mod= e, +use --per-cluster in addition to -a. (system-wide). The output includes t= he +cluster number and the number of online processors on that cluster. This is +useful to gauge the amount of aggregation. The information of cluster ID a= nd +related CPUs can be gotten from /sys/devices/system/cpu/cpuX/topology/clus= ter_{id, cpus}. + --per-cache:: Aggregate counts per cache instance for system-wide mode measurements. By default, the aggregation happens for the cache level at the highest index @@ -396,6 +404,9 @@ Aggregate counts per processor socket for system-wide m= ode measurements. --per-die:: Aggregate counts per processor die for system-wide mode measurements. =20 +--per-cluster:: +Aggregate counts perf processor cluster for system-wide mode measurements. + --per-cache:: Aggregate counts per cache instance for system-wide mode measurements. By default, the aggregation happens for the cache level at the highest index diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 5fe9abc6a524..6bba1a89d030 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -1238,6 +1238,8 @@ static struct option stat_options[] =3D { "aggregate counts per processor socket", AGGR_SOCKET), OPT_SET_UINT(0, "per-die", &stat_config.aggr_mode, "aggregate counts per processor die", AGGR_DIE), + OPT_SET_UINT(0, "per-cluster", &stat_config.aggr_mode, + "aggregate counts per processor cluster", AGGR_CLUSTER), OPT_CALLBACK_OPTARG(0, "per-cache", &stat_config.aggr_mode, &stat_config.= aggr_level, "cache level", "aggregate count at this cache level (Default: LLC)", parse_cache_level), @@ -1428,6 +1430,7 @@ static struct aggr_cpu_id aggr_cpu_id__cache(struct p= erf_cpu cpu, void *data) static const char *const aggr_mode__string[] =3D { [AGGR_CORE] =3D "core", [AGGR_CACHE] =3D "cache", + [AGGR_CLUSTER] =3D "cluster", [AGGR_DIE] =3D "die", [AGGR_GLOBAL] =3D "global", [AGGR_NODE] =3D "node", @@ -1455,6 +1458,12 @@ static struct aggr_cpu_id perf_stat__get_cache_id(st= ruct perf_stat_config *confi return aggr_cpu_id__cache(cpu, /*data=3D*/NULL); } =20 +static struct aggr_cpu_id perf_stat__get_cluster(struct perf_stat_config *= config __maybe_unused, + struct perf_cpu cpu) +{ + return aggr_cpu_id__cluster(cpu, /*data=3D*/NULL); +} + static struct aggr_cpu_id perf_stat__get_core(struct perf_stat_config *con= fig __maybe_unused, struct perf_cpu cpu) { @@ -1507,6 +1516,12 @@ static struct aggr_cpu_id perf_stat__get_die_cached(= struct perf_stat_config *con return perf_stat__get_aggr(config, perf_stat__get_die, cpu); } =20 +static struct aggr_cpu_id perf_stat__get_cluster_cached(struct perf_stat_c= onfig *config, + struct perf_cpu cpu) +{ + return perf_stat__get_aggr(config, perf_stat__get_cluster, cpu); +} + static struct aggr_cpu_id perf_stat__get_cache_id_cached(struct perf_stat_= config *config, struct perf_cpu cpu) { @@ -1544,6 +1559,8 @@ static aggr_cpu_id_get_t aggr_mode__get_aggr(enum agg= r_mode aggr_mode) return aggr_cpu_id__socket; case AGGR_DIE: return aggr_cpu_id__die; + case AGGR_CLUSTER: + return aggr_cpu_id__cluster; case AGGR_CACHE: return aggr_cpu_id__cache; case AGGR_CORE: @@ -1569,6 +1586,8 @@ static aggr_get_id_t aggr_mode__get_id(enum aggr_mode= aggr_mode) return perf_stat__get_socket_cached; case AGGR_DIE: return perf_stat__get_die_cached; + case AGGR_CLUSTER: + return perf_stat__get_cluster_cached; case AGGR_CACHE: return perf_stat__get_cache_id_cached; case AGGR_CORE: @@ -1737,6 +1756,21 @@ static struct aggr_cpu_id perf_env__get_cache_aggr_b= y_cpu(struct perf_cpu cpu, return id; } =20 +static struct aggr_cpu_id perf_env__get_cluster_aggr_by_cpu(struct perf_cp= u cpu, + void *data) +{ + struct perf_env *env =3D data; + struct aggr_cpu_id id =3D aggr_cpu_id__empty(); + + if (cpu.cpu !=3D -1) { + id.socket =3D env->cpu[cpu.cpu].socket_id; + id.die =3D env->cpu[cpu.cpu].die_id; + id.cluster =3D env->cpu[cpu.cpu].cluster_id; + } + + return id; +} + static struct aggr_cpu_id perf_env__get_core_aggr_by_cpu(struct perf_cpu c= pu, void *data) { struct perf_env *env =3D data; @@ -1744,12 +1778,12 @@ static struct aggr_cpu_id perf_env__get_core_aggr_b= y_cpu(struct perf_cpu cpu, vo =20 if (cpu.cpu !=3D -1) { /* - * core_id is relative to socket and die, - * we need a global id. So we set - * socket, die id and core id + * core_id is relative to socket, die and cluster, we need a + * global id. So we set socket, die id, cluster id and core id. */ id.socket =3D env->cpu[cpu.cpu].socket_id; id.die =3D env->cpu[cpu.cpu].die_id; + id.cluster =3D env->cpu[cpu.cpu].cluster_id; id.core =3D env->cpu[cpu.cpu].core_id; } =20 @@ -1805,6 +1839,12 @@ static struct aggr_cpu_id perf_stat__get_die_file(st= ruct perf_stat_config *confi return perf_env__get_die_aggr_by_cpu(cpu, &perf_stat.session->header.env); } =20 +static struct aggr_cpu_id perf_stat__get_cluster_file(struct perf_stat_con= fig *config __maybe_unused, + struct perf_cpu cpu) +{ + return perf_env__get_cluster_aggr_by_cpu(cpu, &perf_stat.session->header.= env); +} + static struct aggr_cpu_id perf_stat__get_cache_file(struct perf_stat_confi= g *config __maybe_unused, struct perf_cpu cpu) { @@ -1842,6 +1882,8 @@ static aggr_cpu_id_get_t aggr_mode__get_aggr_file(enu= m aggr_mode aggr_mode) return perf_env__get_socket_aggr_by_cpu; case AGGR_DIE: return perf_env__get_die_aggr_by_cpu; + case AGGR_CLUSTER: + return perf_env__get_cluster_aggr_by_cpu; case AGGR_CACHE: return perf_env__get_cache_aggr_by_cpu; case AGGR_CORE: @@ -1867,6 +1909,8 @@ static aggr_get_id_t aggr_mode__get_id_file(enum aggr= _mode aggr_mode) return perf_stat__get_socket_file; case AGGR_DIE: return perf_stat__get_die_file; + case AGGR_CLUSTER: + return perf_stat__get_cluster_file; case AGGR_CACHE: return perf_stat__get_cache_file; case AGGR_CORE: @@ -2398,6 +2442,8 @@ static int __cmd_report(int argc, const char **argv) "aggregate counts per processor socket", AGGR_SOCKET), OPT_SET_UINT(0, "per-die", &perf_stat.aggr_mode, "aggregate counts per processor die", AGGR_DIE), + OPT_SET_UINT(0, "per-cluster", &perf_stat.aggr_mode, + "aggregate counts perf processor cluster", AGGR_CLUSTER), OPT_CALLBACK_OPTARG(0, "per-cache", &perf_stat.aggr_mode, &perf_stat.aggr= _level, "cache level", "aggregate count at this cache level (Default: LLC)", diff --git a/tools/perf/tests/shell/lib/perf_json_output_lint.py b/tools/pe= rf/tests/shell/lib/perf_json_output_lint.py index ea55d5ea1ced..abc1fd737782 100644 --- a/tools/perf/tests/shell/lib/perf_json_output_lint.py +++ b/tools/perf/tests/shell/lib/perf_json_output_lint.py @@ -15,6 +15,7 @@ ap.add_argument('--event', action=3D'store_true') ap.add_argument('--per-core', action=3D'store_true') ap.add_argument('--per-thread', action=3D'store_true') ap.add_argument('--per-cache', action=3D'store_true') +ap.add_argument('--per-cluster', action=3D'store_true') ap.add_argument('--per-die', action=3D'store_true') ap.add_argument('--per-node', action=3D'store_true') ap.add_argument('--per-socket', action=3D'store_true') @@ -49,6 +50,7 @@ def check_json_output(expected_items): 'cgroup': lambda x: True, 'cpu': lambda x: isint(x), 'cache': lambda x: True, + 'cluster': lambda x: True, 'die': lambda x: True, 'event': lambda x: True, 'event-runtime': lambda x: isfloat(x), @@ -88,7 +90,7 @@ try: expected_items =3D 7 elif args.interval or args.per_thread or args.system_wide_no_aggr: expected_items =3D 8 - elif args.per_core or args.per_socket or args.per_node or args.per_die o= r args.per_cache: + elif args.per_core or args.per_socket or args.per_node or args.per_die o= r args.per_cluster or args.per_cache: expected_items =3D 9 else: # If no option is specified, don't check the number of items. diff --git a/tools/perf/tests/shell/lib/stat_output.sh b/tools/perf/tests/s= hell/lib/stat_output.sh index 3cc158a64326..c81d6a9f7983 100644 --- a/tools/perf/tests/shell/lib/stat_output.sh +++ b/tools/perf/tests/shell/lib/stat_output.sh @@ -97,6 +97,18 @@ check_per_cache_instance() echo "[Success]" } =20 +check_per_cluster() +{ + echo -n "Checking $1 output: per cluster " + if ParanoidAndNotRoot 0 + then + echo "[Skip] paranoid and not root" + return + fi + perf stat --per-cluster -a $2 true + echo "[Success]" +} + check_per_die() { echo -n "Checking $1 output: per die " diff --git a/tools/perf/tests/shell/stat+csv_output.sh b/tools/perf/tests/s= hell/stat+csv_output.sh index f1818fa6d9ce..fc2d8cc6e5e0 100755 --- a/tools/perf/tests/shell/stat+csv_output.sh +++ b/tools/perf/tests/shell/stat+csv_output.sh @@ -42,6 +42,7 @@ function commachecker() ;; "--per-socket") exp=3D8 ;; "--per-node") exp=3D8 ;; "--per-die") exp=3D8 + ;; "--per-cluster") exp=3D8 ;; "--per-cache") exp=3D8 esac =20 @@ -79,6 +80,7 @@ then check_system_wide_no_aggr "CSV" "$perf_cmd" check_per_core "CSV" "$perf_cmd" check_per_cache_instance "CSV" "$perf_cmd" + check_per_cluster "CSV" "$perf_cmd" check_per_die "CSV" "$perf_cmd" check_per_socket "CSV" "$perf_cmd" else diff --git a/tools/perf/tests/shell/stat+json_output.sh b/tools/perf/tests/= shell/stat+json_output.sh index 3bc900533a5d..2b9c6212dffc 100755 --- a/tools/perf/tests/shell/stat+json_output.sh +++ b/tools/perf/tests/shell/stat+json_output.sh @@ -122,6 +122,18 @@ check_per_cache_instance() echo "[Success]" } =20 +check_per_cluster() +{ + echo -n "Checking json output: per cluster " + if ParanoidAndNotRoot 0 + then + echo "[Skip] paranoia and not root" + return + fi + perf stat -j --per-cluster -a true 2>&1 | $PYTHON $pythonchecker --per-cl= uster + echo "[Success]" +} + check_per_die() { echo -n "Checking json output: per die " @@ -200,6 +212,7 @@ then check_system_wide_no_aggr check_per_core check_per_cache_instance + check_per_cluster check_per_die check_per_socket else diff --git a/tools/perf/tests/shell/stat+std_output.sh b/tools/perf/tests/s= hell/stat+std_output.sh index 4fcdd1a9142c..16f61e86afc5 100755 --- a/tools/perf/tests/shell/stat+std_output.sh +++ b/tools/perf/tests/shell/stat+std_output.sh @@ -40,6 +40,7 @@ function commachecker() ;; "--per-node") prefix=3D3 ;; "--per-die") prefix=3D3 ;; "--per-cache") prefix=3D3 + ;; "--per-cluster") prefix=3D3 esac =20 while read line @@ -99,6 +100,7 @@ then check_system_wide_no_aggr "STD" "$perf_cmd" check_per_core "STD" "$perf_cmd" check_per_cache_instance "STD" "$perf_cmd" + check_per_cluster "STD" "$perf_cmd" check_per_die "STD" "$perf_cmd" check_per_socket "STD" "$perf_cmd" else diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c index 0581ee0fa5f2..356e30c42cd8 100644 --- a/tools/perf/util/cpumap.c +++ b/tools/perf/util/cpumap.c @@ -222,6 +222,8 @@ static int aggr_cpu_id__cmp(const void *a_pointer, cons= t void *b_pointer) return a->socket - b->socket; else if (a->die !=3D b->die) return a->die - b->die; + else if (a->cluster !=3D b->cluster) + return a->cluster - b->cluster; else if (a->cache_lvl !=3D b->cache_lvl) return a->cache_lvl - b->cache_lvl; else if (a->cache !=3D b->cache) @@ -309,6 +311,30 @@ struct aggr_cpu_id aggr_cpu_id__die(struct perf_cpu cp= u, void *data) return id; } =20 +int cpu__get_cluster_id(struct perf_cpu cpu) +{ + int value, ret =3D cpu__get_topology_int(cpu.cpu, "cluster_id", &value); + + return ret ?: value; +} + +struct aggr_cpu_id aggr_cpu_id__cluster(struct perf_cpu cpu, void *data) +{ + int cluster =3D cpu__get_cluster_id(cpu); + struct aggr_cpu_id id; + + /* There is no cluster_id on legacy system. */ + if (cluster =3D=3D -1) + cluster =3D 0; + + id =3D aggr_cpu_id__die(cpu, data); + if (aggr_cpu_id__is_empty(&id)) + return id; + + id.cluster =3D cluster; + return id; +} + int cpu__get_core_id(struct perf_cpu cpu) { int value, ret =3D cpu__get_topology_int(cpu.cpu, "core_id", &value); @@ -320,8 +346,8 @@ struct aggr_cpu_id aggr_cpu_id__core(struct perf_cpu cp= u, void *data) struct aggr_cpu_id id; int core =3D cpu__get_core_id(cpu); =20 - /* aggr_cpu_id__die returns a struct with socket and die set. */ - id =3D aggr_cpu_id__die(cpu, data); + /* aggr_cpu_id__die returns a struct with socket die, and cluster set. */ + id =3D aggr_cpu_id__cluster(cpu, data); if (aggr_cpu_id__is_empty(&id)) return id; =20 @@ -683,6 +709,7 @@ bool aggr_cpu_id__equal(const struct aggr_cpu_id *a, co= nst struct aggr_cpu_id *b a->node =3D=3D b->node && a->socket =3D=3D b->socket && a->die =3D=3D b->die && + a->cluster =3D=3D b->cluster && a->cache_lvl =3D=3D b->cache_lvl && a->cache =3D=3D b->cache && a->core =3D=3D b->core && @@ -695,6 +722,7 @@ bool aggr_cpu_id__is_empty(const struct aggr_cpu_id *a) a->node =3D=3D -1 && a->socket =3D=3D -1 && a->die =3D=3D -1 && + a->cluster =3D=3D -1 && a->cache_lvl =3D=3D -1 && a->cache =3D=3D -1 && a->core =3D=3D -1 && @@ -708,6 +736,7 @@ struct aggr_cpu_id aggr_cpu_id__empty(void) .node =3D -1, .socket =3D -1, .die =3D -1, + .cluster =3D -1, .cache_lvl =3D -1, .cache =3D -1, .core =3D -1, diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h index 9df2aeb34d3d..26cf76c693f5 100644 --- a/tools/perf/util/cpumap.h +++ b/tools/perf/util/cpumap.h @@ -20,6 +20,8 @@ struct aggr_cpu_id { int socket; /** The die id as read from /sys/devices/system/cpu/cpuX/topology/die_id.= */ int die; + /** The cluster id as read from /sys/devices/system/cpu/cpuX/topology/clu= ster_id */ + int cluster; /** The cache level as read from /sys/devices/system/cpu/cpuX/cache/index= Y/level */ int cache_lvl; /** @@ -86,6 +88,11 @@ int cpu__get_socket_id(struct perf_cpu cpu); * /sys/devices/system/cpu/cpuX/topology/die_id for the given CPU. */ int cpu__get_die_id(struct perf_cpu cpu); +/** + * cpu__get_cluster_id - Returns the cluster id as read from + * /sys/devices/system/cpu/cpuX/topology/cluster_id for the given CPU + */ +int cpu__get_cluster_id(struct perf_cpu cpu); /** * cpu__get_core_id - Returns the core id as read from * /sys/devices/system/cpu/cpuX/topology/core_id for the given CPU. @@ -127,9 +134,15 @@ struct aggr_cpu_id aggr_cpu_id__socket(struct perf_cpu= cpu, void *data); */ struct aggr_cpu_id aggr_cpu_id__die(struct perf_cpu cpu, void *data); /** - * aggr_cpu_id__core - Create an aggr_cpu_id with the core, die and socket - * populated with the core, die and socket for cpu. The function signature= is - * compatible with aggr_cpu_id_get_t. + * aggr_cpu_id__cluster - Create an aggr_cpu_id with cluster, die and sock= et + * populated with the cluster, die and socket for cpu. The function signat= ure + * is compatible with aggr_cpu_id_get_t. + */ +struct aggr_cpu_id aggr_cpu_id__cluster(struct perf_cpu cpu, void *data); +/** + * aggr_cpu_id__core - Create an aggr_cpu_id with the core, cluster, die a= nd + * socket populated with the core, die and socket for cpu. The function + * signature is compatible with aggr_cpu_id_get_t. */ struct aggr_cpu_id aggr_cpu_id__core(struct perf_cpu cpu, void *data); /** diff --git a/tools/perf/util/env.h b/tools/perf/util/env.h index 7c527e65c186..2a2c37cc40b7 100644 --- a/tools/perf/util/env.h +++ b/tools/perf/util/env.h @@ -12,6 +12,7 @@ struct perf_cpu_map; struct cpu_topology_map { int socket_id; int die_id; + int cluster_id; int core_id; }; =20 diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c index 8c61f8627ebc..4dfe7d9517a9 100644 --- a/tools/perf/util/stat-display.c +++ b/tools/perf/util/stat-display.c @@ -201,6 +201,9 @@ static void print_aggr_id_std(struct perf_stat_config *= config, snprintf(buf, sizeof(buf), "S%d-D%d-L%d-ID%d", id.socket, id.die, id.cache_lvl, id.cache); break; + case AGGR_CLUSTER: + snprintf(buf, sizeof(buf), "S%d-D%d-CLS%d", id.socket, id.die, id.cluste= r); + break; case AGGR_DIE: snprintf(buf, sizeof(buf), "S%d-D%d", id.socket, id.die); break; @@ -251,6 +254,10 @@ static void print_aggr_id_csv(struct perf_stat_config = *config, fprintf(config->output, "S%d-D%d-L%d-ID%d%s%d%s", id.socket, id.die, id.cache_lvl, id.cache, sep, aggr_nr, sep); break; + case AGGR_CLUSTER: + fprintf(config->output, "S%d-D%d-CLS%d%s%d%s", + id.socket, id.die, id.cluster, sep, aggr_nr, sep); + break; case AGGR_DIE: fprintf(output, "S%d-D%d%s%d%s", id.socket, id.die, sep, aggr_nr, sep); @@ -300,6 +307,10 @@ static void print_aggr_id_json(struct perf_stat_config= *config, fprintf(output, "\"cache\" : \"S%d-D%d-L%d-ID%d\", \"aggregate-number\" = : %d, ", id.socket, id.die, id.cache_lvl, id.cache, aggr_nr); break; + case AGGR_CLUSTER: + fprintf(output, "\"cluster\" : \"S%d-D%d-CLS%d\", \"aggregate-number\" := %d, ", + id.socket, id.die, id.cluster, aggr_nr); + break; case AGGR_DIE: fprintf(output, "\"die\" : \"S%d-D%d\", \"aggregate-number\" : %d, ", id.socket, id.die, aggr_nr); @@ -1248,6 +1259,7 @@ static void print_header_interval_std(struct perf_sta= t_config *config, case AGGR_NODE: case AGGR_SOCKET: case AGGR_DIE: + case AGGR_CLUSTER: case AGGR_CACHE: case AGGR_CORE: fprintf(output, "#%*s %-*s cpus", @@ -1550,6 +1562,7 @@ void evlist__print_counters(struct evlist *evlist, st= ruct perf_stat_config *conf switch (config->aggr_mode) { case AGGR_CORE: case AGGR_CACHE: + case AGGR_CLUSTER: case AGGR_DIE: case AGGR_SOCKET: case AGGR_NODE: diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h index 4357ba114822..d6e5c8787ba2 100644 --- a/tools/perf/util/stat.h +++ b/tools/perf/util/stat.h @@ -48,6 +48,7 @@ enum aggr_mode { AGGR_GLOBAL, AGGR_SOCKET, AGGR_DIE, + AGGR_CLUSTER, AGGR_CACHE, AGGR_CORE, AGGR_THREAD, --=20 2.24.0