summaryrefslogtreecommitdiffstats
path: root/tools/perf/builtin-stat.c
diff options
context:
space:
mode:
Diffstat (limited to 'tools/perf/builtin-stat.c')
-rw-r--r--tools/perf/builtin-stat.c97
1 files changed, 86 insertions, 11 deletions
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 1ae66f09dc7d..b55a534b4de0 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -43,7 +43,6 @@
#include "perf.h"
#include "builtin.h"
#include "util/cgroup.h"
-#include "util/util.h"
#include <subcmd/parse-options.h>
#include "util/parse-events.h"
#include "util/pmu.h"
@@ -67,6 +66,7 @@
#include "asm/bug.h"
#include <linux/time64.h>
+#include <linux/zalloc.h>
#include <api/fs/fs.h>
#include <errno.h>
#include <signal.h>
@@ -82,7 +82,7 @@
#include <sys/time.h>
#include <sys/resource.h>
-#include "sane_ctype.h"
+#include <linux/ctype.h>
#define DEFAULT_SEPARATOR " "
#define FREEZE_ON_SMI_PATH "devices/cpu/freeze_on_smi"
@@ -776,6 +776,8 @@ 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,
@@ -800,6 +802,12 @@ 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)
{
@@ -840,6 +848,12 @@ 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)
{
@@ -870,6 +884,13 @@ 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");
@@ -935,21 +956,55 @@ 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_id = -1, cpu = perf_env__get_cpu(env, map, idx);
+
+ if (cpu != -1) {
+ /*
+ * Encode socket in bit range 15:8
+ * die_id is relative to socket,
+ * we need a global id. So we combine
+ * socket + die id
+ */
+ if (WARN_ONCE(env->cpu[cpu].socket_id >> 8, "The socket id number is too big.\n"))
+ return -1;
+
+ if (WARN_ONCE(env->cpu[cpu].die_id >> 8, "The die id number is too big.\n"))
+ return -1;
+
+ die_id = (env->cpu[cpu].socket_id << 8) | (env->cpu[cpu].die_id & 0xff);
+ }
+
+ return die_id;
+}
+
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 bit range 31:24
+ * encode die id in bit range 23:16
+ * 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);
+ if (WARN_ONCE(env->cpu[cpu].socket_id >> 8, "The socket id number is too big.\n"))
+ return -1;
+
+ if (WARN_ONCE(env->cpu[cpu].die_id >> 8, "The die id number is too big.\n"))
+ return -1;
+
+ if (WARN_ONCE(env->cpu[cpu].core_id >> 16, "The core id number is too big.\n"))
+ return -1;
+
+ core = (env->cpu[cpu].socket_id << 24) |
+ (env->cpu[cpu].die_id << 16) |
+ (env->cpu[cpu].core_id & 0xffff);
}
return core;
@@ -961,6 +1016,12 @@ 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)
{
@@ -972,6 +1033,11 @@ 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)
@@ -991,6 +1057,13 @@ 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");
@@ -1276,8 +1349,8 @@ static int add_default_attributes(void)
fprintf(stderr,
"Cannot set up top down events %s: %d\n",
str, err);
- free(str);
parse_events_print_error(&errinfo, str);
+ free(str);
return -1;
}
} else {
@@ -1513,7 +1586,7 @@ static void runtime_stat_delete(struct perf_stat_config *config)
for (i = 0; i < config->stats_num; i++)
runtime_stat__exit(&config->stats[i]);
- free(config->stats);
+ zfree(&config->stats);
}
static const char * const stat_report_usage[] = {
@@ -1541,6 +1614,8 @@ 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,
@@ -1928,7 +2003,7 @@ int cmd_stat(int argc, const char **argv)
perf_stat__exit_aggr_mode();
perf_evlist__free_stats(evsel_list);
out:
- free(stat_config.walltime_run);
+ zfree(&stat_config.walltime_run);
if (smi_cost && smi_reset)
sysfs__write_int(FREEZE_ON_SMI_PATH, 0);
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2025-05-01 08:20:56 +0200