From: kan.liang@linux.intel.com
To: peterz@infradead.org, tglx@linutronix.de, acme@kernel.org,
mingo@redhat.com, x86@kernel.org, linux-kernel@vger.kernel.org
Cc: len.brown@intel.com, jolsa@redhat.com, namhyung@kernel.org,
eranian@google.com, ak@linux.intel.com,
Kan Liang <kan.liang@linux.intel.com>
Subject: [PATCH 10/10] perf stat: Support per-die aggregation
Date: Tue, 19 Feb 2019 12:00:11 -0800 [thread overview]
Message-ID: <1550606411-5313-11-git-send-email-kan.liang@linux.intel.com> (raw)
In-Reply-To: <1550606411-5313-1-git-send-email-kan.liang@linux.intel.com>
From: Kan Liang <kan.liang@linux.intel.com>
It is useful to aggregate counts per processor die.
Introduce a new option "--per-die" to support per-die aggregation.
The global id for each core has been changed to socket + die id + core
id. The global id for each die is socket + die id.
Add die information for per-core aggregation. The output of per-core
aggregation will be changed from "S0-C0" to "S0-D0-C0". Any scripts
which rely on the output format of per-core aggregation probably be
broken.
For perf stat record/report, there is no die information when processing
the old perf.data. The per-die result will be the same as per-socket.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
---
tools/perf/Documentation/perf-stat.txt | 10 +++++
tools/perf/builtin-stat.c | 73 +++++++++++++++++++++++++++++++---
tools/perf/util/cpumap.c | 48 +++++++++++++++++++---
tools/perf/util/cpumap.h | 9 ++++-
tools/perf/util/stat-display.c | 24 ++++++++++-
tools/perf/util/stat-shadow.c | 1 +
tools/perf/util/stat.c | 1 +
tools/perf/util/stat.h | 1 +
8 files changed, 153 insertions(+), 14 deletions(-)
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 4bc2085..4fd68ac 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -197,6 +197,13 @@ use --per-socket in addition to -a. (system-wide). The output includes the
socket number and the number of online processors on that socket. This is
useful to gauge the amount of aggregation.
+--per-die::
+Aggregate counts per processor die for system-wide mode measurements. This
+is a useful mode to detect imbalance between dies. To enable this mode,
+use --per-die in addition to -a. (system-wide). The output includes the
+die number and the number of online processors on that die. This is
+useful to gauge the amount of aggregation.
+
--per-core::
Aggregate counts per physical processor for system-wide mode measurements. This
is a useful mode to detect imbalance between physical cores. To enable this mode,
@@ -236,6 +243,9 @@ Input file name.
--per-socket::
Aggregate counts per processor socket for system-wide mode measurements.
+--per-die::
+Aggregate counts per processor die for system-wide mode measurements.
+
--per-core::
Aggregate counts per physical processor for system-wide mode measurements.
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 63a3afc..92c6967 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -772,6 +772,8 @@ static struct option stat_options[] = {
"stop workload and print counts after a timeout period in ms (>= 10ms)"),
OPT_SET_UINT(0, "per-socket", &stat_config.aggr_mode,
"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-core", &stat_config.aggr_mode,
"aggregate counts per physical processor core", AGGR_CORE),
OPT_SET_UINT(0, "per-thread", &stat_config.aggr_mode,
@@ -796,6 +798,12 @@ static int perf_stat__get_socket(struct perf_stat_config *config __maybe_unused,
return cpu_map__get_socket(map, cpu, NULL);
}
+static int perf_stat__get_die(struct perf_stat_config *config __maybe_unused,
+ struct cpu_map *map, int cpu)
+{
+ return cpu_map__get_die(map, cpu, NULL);
+}
+
static int perf_stat__get_core(struct perf_stat_config *config __maybe_unused,
struct cpu_map *map, int cpu)
{
@@ -836,6 +844,12 @@ static int perf_stat__get_socket_cached(struct perf_stat_config *config,
return perf_stat__get_aggr(config, perf_stat__get_socket, map, idx);
}
+static int perf_stat__get_die_cached(struct perf_stat_config *config,
+ struct cpu_map *map, int idx)
+{
+ return perf_stat__get_aggr(config, perf_stat__get_die, map, idx);
+}
+
static int perf_stat__get_core_cached(struct perf_stat_config *config,
struct cpu_map *map, int idx)
{
@@ -854,6 +868,13 @@ static int perf_stat_init_aggr_mode(void)
}
stat_config.aggr_get_id = perf_stat__get_socket_cached;
break;
+ case AGGR_DIE:
+ if (cpu_map__build_die_map(evsel_list->cpus, &stat_config.aggr_map)) {
+ perror("cannot build die map");
+ return -1;
+ }
+ stat_config.aggr_get_id = perf_stat__get_die_cached;
+ break;
case AGGR_CORE:
if (cpu_map__build_core_map(evsel_list->cpus, &stat_config.aggr_map)) {
perror("cannot build core map");
@@ -910,21 +931,41 @@ static int perf_env__get_socket(struct cpu_map *map, int idx, void *data)
return cpu == -1 ? -1 : env->cpu[cpu].socket_id;
}
+static int perf_env__get_die(struct cpu_map *map, int idx, void *data)
+{
+ struct perf_env *env = data;
+ int die = -1, cpu = perf_env__get_cpu(env, map, idx);
+
+ if (cpu != -1) {
+ /*
+ * Encode socket in upper 8 bits
+ * die_id is relative to socket,
+ * we need a global id. So we combine
+ * socket + die id
+ */
+ die = (env->cpu[cpu].socket_id << 8) |
+ (env->cpu[cpu].die_id & 0xff);
+ }
+
+ return die;
+}
+
static int perf_env__get_core(struct cpu_map *map, int idx, void *data)
{
struct perf_env *env = data;
int core = -1, cpu = perf_env__get_cpu(env, map, idx);
if (cpu != -1) {
- int socket_id = env->cpu[cpu].socket_id;
-
/*
- * Encode socket in upper 16 bits
- * core_id is relative to socket, and
+ * Encode socket in upper 24 bits
+ * encode die id in upper 16 bits
+ * core_id is relative to socket and die,
* we need a global id. So we combine
- * socket + core id.
+ * socket + die id + core id
*/
- core = (socket_id << 16) | (env->cpu[cpu].core_id & 0xffff);
+ core = (env->cpu[cpu].socket_id << 24) |
+ (env->cpu[cpu].die_id << 16) |
+ (env->cpu[cpu].core_id & 0xffff);
}
return core;
@@ -936,6 +977,12 @@ static int perf_env__build_socket_map(struct perf_env *env, struct cpu_map *cpus
return cpu_map__build_map(cpus, sockp, perf_env__get_socket, env);
}
+static int perf_env__build_die_map(struct perf_env *env, struct cpu_map *cpus,
+ struct cpu_map **diep)
+{
+ return cpu_map__build_map(cpus, diep, perf_env__get_die, env);
+}
+
static int perf_env__build_core_map(struct perf_env *env, struct cpu_map *cpus,
struct cpu_map **corep)
{
@@ -947,6 +994,11 @@ static int perf_stat__get_socket_file(struct perf_stat_config *config __maybe_un
{
return perf_env__get_socket(map, idx, &perf_stat.session->header.env);
}
+static int perf_stat__get_die_file(struct perf_stat_config *config __maybe_unused,
+ struct cpu_map *map, int idx)
+{
+ return perf_env__get_die(map, idx, &perf_stat.session->header.env);
+}
static int perf_stat__get_core_file(struct perf_stat_config *config __maybe_unused,
struct cpu_map *map, int idx)
@@ -966,6 +1018,13 @@ static int perf_stat_init_aggr_mode_file(struct perf_stat *st)
}
stat_config.aggr_get_id = perf_stat__get_socket_file;
break;
+ case AGGR_DIE:
+ if (perf_env__build_die_map(env, evsel_list->cpus, &stat_config.aggr_map)) {
+ perror("cannot build die map");
+ return -1;
+ }
+ stat_config.aggr_get_id = perf_stat__get_die_file;
+ break;
case AGGR_CORE:
if (perf_env__build_core_map(env, evsel_list->cpus, &stat_config.aggr_map)) {
perror("cannot build core map");
@@ -1515,6 +1574,8 @@ static int __cmd_report(int argc, const char **argv)
OPT_STRING('i', "input", &input_name, "file", "input file name"),
OPT_SET_UINT(0, "per-socket", &perf_stat.aggr_mode,
"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-core", &perf_stat.aggr_mode,
"aggregate counts per physical processor core", AGGR_CORE),
OPT_SET_UINT('A', "no-aggr", &perf_stat.aggr_mode,
diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index b84fcd4..f6c7aa1 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -380,6 +380,33 @@ int cpu_map__get_die_id(int cpu)
return ret ?: value;
}
+int cpu_map__get_die(struct cpu_map *map, int idx, void *data)
+{
+ int cpu, die, s;
+
+ if (idx > map->nr)
+ return -1;
+
+ cpu = map->map[idx];
+
+ die = cpu_map__get_die_id(cpu);
+ /* There is no die_id on legacy system. */
+ if (die == -1)
+ die = 0;
+
+ s = cpu_map__get_socket(map, idx, data);
+ if (s == -1)
+ return -1;
+
+ /*
+ * encode socket in upper 8 bits
+ * die_id is relative to socket, and
+ * we need a global id. So we combine
+ * socket + die id
+ */
+ return (s << 8) | (die & 0xff);
+}
+
int cpu_map__get_core_id(int cpu)
{
int value, ret = cpu__get_topology_int(cpu, "core_id", &value);
@@ -388,7 +415,7 @@ int cpu_map__get_core_id(int cpu)
int cpu_map__get_core(struct cpu_map *map, int idx, void *data)
{
- int cpu, s;
+ int cpu, s, die;
if (idx > map->nr)
return -1;
@@ -397,17 +424,23 @@ int cpu_map__get_core(struct cpu_map *map, int idx, void *data)
cpu = cpu_map__get_core_id(cpu);
+ die = cpu_map__get_die(map, idx, data);
+ /* There is no die_id on legacy system. */
+ if (die == -1)
+ die = 0;
+
s = cpu_map__get_socket(map, idx, data);
if (s == -1)
return -1;
/*
- * encode socket in upper 16 bits
- * core_id is relative to socket, and
+ * encode socket in upper 24 bits
+ * encode die id in upper 16 bits
+ * core_id is relative to socket and die,
* we need a global id. So we combine
- * socket+ core id
+ * socket + die id + core id
*/
- return (s << 16) | (cpu & 0xffff);
+ return (s << 24) | (die << 16) | (cpu & 0xffff);
}
int cpu_map__build_socket_map(struct cpu_map *cpus, struct cpu_map **sockp)
@@ -415,6 +448,11 @@ int cpu_map__build_socket_map(struct cpu_map *cpus, struct cpu_map **sockp)
return cpu_map__build_map(cpus, sockp, cpu_map__get_socket, NULL);
}
+int cpu_map__build_die_map(struct cpu_map *cpus, struct cpu_map **diep)
+{
+ return cpu_map__build_map(cpus, diep, cpu_map__get_die, NULL);
+}
+
int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep)
{
return cpu_map__build_map(cpus, corep, cpu_map__get_core, NULL);
diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h
index 4d66f67..3442b7a 100644
--- a/tools/perf/util/cpumap.h
+++ b/tools/perf/util/cpumap.h
@@ -26,9 +26,11 @@ size_t cpu_map__fprintf(struct cpu_map *map, FILE *fp);
int cpu_map__get_socket_id(int cpu);
int cpu_map__get_socket(struct cpu_map *map, int idx, void *data);
int cpu_map__get_die_id(int cpu);
+int cpu_map__get_die(struct cpu_map *map, int idx, void *data);
int cpu_map__get_core_id(int cpu);
int cpu_map__get_core(struct cpu_map *map, int idx, void *data);
int cpu_map__build_socket_map(struct cpu_map *cpus, struct cpu_map **sockp);
+int cpu_map__build_die_map(struct cpu_map *cpus, struct cpu_map **diep);
int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep);
struct cpu_map *cpu_map__get(struct cpu_map *map);
@@ -43,7 +45,12 @@ static inline int cpu_map__socket(struct cpu_map *sock, int s)
static inline int cpu_map__id_to_socket(int id)
{
- return id >> 16;
+ return id >> 24;
+}
+
+static inline int cpu_map__id_to_die(int id)
+{
+ return (id >> 16) & 0xff;
}
static inline int cpu_map__id_to_cpu(int id)
diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c
index 665ee37..05216a2 100644
--- a/tools/perf/util/stat-display.c
+++ b/tools/perf/util/stat-display.c
@@ -73,8 +73,9 @@ static void aggr_printout(struct perf_stat_config *config,
{
switch (config->aggr_mode) {
case AGGR_CORE:
- fprintf(config->output, "S%d-C%*d%s%*d%s",
+ fprintf(config->output, "S%d-D%d-C%*d%s%*d%s",
cpu_map__id_to_socket(id),
+ cpu_map__id_to_die(id),
config->csv_output ? 0 : -8,
cpu_map__id_to_cpu(id),
config->csv_sep,
@@ -82,6 +83,16 @@ static void aggr_printout(struct perf_stat_config *config,
nr,
config->csv_sep);
break;
+ case AGGR_DIE:
+ fprintf(config->output, "S%d-D%*d%s%*d%s",
+ cpu_map__id_to_socket(id << 16),
+ config->csv_output ? 0 : -8,
+ cpu_map__id_to_die(id << 16),
+ config->csv_sep,
+ config->csv_output ? 0 : 4,
+ nr,
+ config->csv_sep);
+ break;
case AGGR_SOCKET:
fprintf(config->output, "S%*d%s%*d%s",
config->csv_output ? 0 : -5,
@@ -403,6 +414,7 @@ static void printout(struct perf_stat_config *config, int id, int nr,
[AGGR_THREAD] = 1,
[AGGR_NONE] = 1,
[AGGR_SOCKET] = 2,
+ [AGGR_DIE] = 2,
[AGGR_CORE] = 2,
};
@@ -866,7 +878,8 @@ static void print_no_aggr_metric(struct perf_stat_config *config,
}
static int aggr_header_lens[] = {
- [AGGR_CORE] = 18,
+ [AGGR_CORE] = 24,
+ [AGGR_DIE] = 18,
[AGGR_SOCKET] = 12,
[AGGR_NONE] = 6,
[AGGR_THREAD] = 24,
@@ -875,6 +888,7 @@ static int aggr_header_lens[] = {
static const char *aggr_header_csv[] = {
[AGGR_CORE] = "core,cpus,",
+ [AGGR_DIE] = "die,cpus",
[AGGR_SOCKET] = "socket,cpus",
[AGGR_NONE] = "cpu,",
[AGGR_THREAD] = "comm-pid,",
@@ -943,6 +957,11 @@ static void print_interval(struct perf_stat_config *config,
if (!metric_only)
fprintf(output, " counts %*s events\n", unit_width, "unit");
break;
+ case AGGR_DIE:
+ fprintf(output, "# time die cpus");
+ if (!metric_only)
+ fprintf(output, " counts %*s events\n", unit_width, "unit");
+ break;
case AGGR_CORE:
fprintf(output, "# time core cpus");
if (!metric_only)
@@ -1130,6 +1149,7 @@ perf_evlist__print_counters(struct perf_evlist *evlist,
switch (config->aggr_mode) {
case AGGR_CORE:
+ case AGGR_DIE:
case AGGR_SOCKET:
print_aggr(config, evlist, prefix);
break;
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index 3c22c58..4d34dc8 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -12,6 +12,7 @@
/*
* AGGR_GLOBAL: Use CPU 0
* AGGR_SOCKET: Use first CPU of socket
+ * AGGR_DIE: Use first CPU of die
* AGGR_CORE: Use first CPU of core
* AGGR_NONE: Use matching CPU
* AGGR_THREAD: Not supported?
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
index 4d40515..6bad12f 100644
--- a/tools/perf/util/stat.c
+++ b/tools/perf/util/stat.c
@@ -272,6 +272,7 @@ process_counter_values(struct perf_stat_config *config, struct perf_evsel *evsel
switch (config->aggr_mode) {
case AGGR_THREAD:
case AGGR_CORE:
+ case AGGR_DIE:
case AGGR_SOCKET:
case AGGR_NONE:
if (!evsel->snapshot)
diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h
index 2f9c915..7032dd1 100644
--- a/tools/perf/util/stat.h
+++ b/tools/perf/util/stat.h
@@ -44,6 +44,7 @@ enum aggr_mode {
AGGR_NONE,
AGGR_GLOBAL,
AGGR_SOCKET,
+ AGGR_DIE,
AGGR_CORE,
AGGR_THREAD,
AGGR_UNSET,
--
2.7.4
next prev parent reply other threads:[~2019-02-19 20:01 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-19 20:00 [PATCH 00/10] perf: Multi-die/package support kan.liang
2019-02-19 20:00 ` [PATCH 01/10] perf/x86/intel: Introduce a concept "domain" as the scope of counters kan.liang
2019-02-20 11:12 ` Peter Zijlstra
2019-02-20 14:36 ` Liang, Kan
2019-03-05 20:32 ` Liang, Kan
2019-02-19 20:00 ` [PATCH 02/10] perf/x86/intel/cstate: Apply "domain" for cstate kan.liang
2019-02-19 20:00 ` [PATCH 03/10] perf/x86/intel/uncore: Apply "domain" for uncore kan.liang
2019-02-19 20:00 ` [PATCH 04/10] perf/x86/intel/rapl: Apply "domain" for RAPL kan.liang
2019-02-19 20:00 ` [PATCH 05/10] perf/x86/intel/domain: Add new domain type for die kan.liang
2019-02-19 20:00 ` [PATCH 06/10] perf/x86/intel/cstate: Support die scope counters on CLX-AP kan.liang
2019-02-19 20:00 ` [PATCH 07/10] perf/x86/intel/uncore: " kan.liang
2019-02-19 20:00 ` [PATCH 08/10] perf/x86/intel/rapl: " kan.liang
2019-02-19 20:00 ` [PATCH 09/10] perf header: Add die information in cpu topology kan.liang
2019-02-19 20:00 ` kan.liang [this message]
2019-02-20 10:15 ` [PATCH 00/10] perf: Multi-die/package support Peter Zijlstra
2019-02-20 12:46 ` Jiri Olsa
2019-02-20 13:24 ` Peter Zijlstra
2019-02-20 13:32 ` Jiri Olsa
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1550606411-5313-11-git-send-email-kan.liang@linux.intel.com \
--to=kan.liang@linux.intel.com \
--cc=acme@kernel.org \
--cc=ak@linux.intel.com \
--cc=eranian@google.com \
--cc=jolsa@redhat.com \
--cc=len.brown@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).