6 files changed, 221 insertions, 131 deletions

diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index dc5c5e6fc502..87d3eca9b872 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -95,6 +95,23 @@ struct perf_cpu_map *perf_cpu_map__empty_new(int nr)
 	return cpus;
 }
 
+struct cpu_aggr_map *cpu_aggr_map__empty_new(int nr)
+{
+	struct cpu_aggr_map *cpus = malloc(sizeof(*cpus) + sizeof(struct aggr_cpu_id) * nr);
+
+	if (cpus != NULL) {
+		int i;
+
+		cpus->nr = nr;
+		for (i = 0; i < nr; i++)
+			cpus->map[i] = cpu_map__empty_aggr_cpu_id();
+
+		refcount_set(&cpus->refcnt, 1);
+	}
+
+	return cpus;
+}
+
 static int cpu__get_topology_int(int cpu, const char *name, int *value)
 {
 	char path[PATH_MAX];
@@ -111,40 +128,57 @@ int cpu_map__get_socket_id(int cpu)
 	return ret ?: value;
 }
 
-int cpu_map__get_socket(struct perf_cpu_map *map, int idx, void *data __maybe_unused)
+struct aggr_cpu_id cpu_map__get_socket(struct perf_cpu_map *map, int idx,
+					void *data __maybe_unused)
 {
 	int cpu;
+	struct aggr_cpu_id id = cpu_map__empty_aggr_cpu_id();
 
 	if (idx > map->nr)
-		return -1;
+		return id;
 
 	cpu = map->map[idx];
 
-	return cpu_map__get_socket_id(cpu);
+	id.socket = cpu_map__get_socket_id(cpu);
+	return id;
 }
 
-static int cmp_ids(const void *a, const void *b)
+static int cmp_aggr_cpu_id(const void *a_pointer, const void *b_pointer)
 {
-	return *(int *)a - *(int *)b;
+	struct aggr_cpu_id *a = (struct aggr_cpu_id *)a_pointer;
+	struct aggr_cpu_id *b = (struct aggr_cpu_id *)b_pointer;
+
+	if (a->node != b->node)
+		return a->node - b->node;
+	else if (a->socket != b->socket)
+		return a->socket - b->socket;
+	else if (a->die != b->die)
+		return a->die - b->die;
+	else if (a->core != b->core)
+		return a->core - b->core;
+	else
+		return a->thread - b->thread;
 }
 
-int cpu_map__build_map(struct perf_cpu_map *cpus, struct perf_cpu_map **res,
-		       int (*f)(struct perf_cpu_map *map, int cpu, void *data),
+int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,
+		       struct aggr_cpu_id (*f)(struct perf_cpu_map *map, int cpu, void *data),
 		       void *data)
 {
-	struct perf_cpu_map *c;
 	int nr = cpus->nr;
-	int cpu, s1, s2;
+	struct cpu_aggr_map *c = cpu_aggr_map__empty_new(nr);
+	int cpu, s2;
+	struct aggr_cpu_id s1;
 
-	/* allocate as much as possible */
-	c = calloc(1, sizeof(*c) + nr * sizeof(int));
 	if (!c)
 		return -1;
 
+	/* Reset size as it may only be partially filled */
+	c->nr = 0;
+
 	for (cpu = 0; cpu < nr; cpu++) {
 		s1 = f(cpus, cpu, data);
 		for (s2 = 0; s2 < c->nr; s2++) {
-			if (s1 == c->map[s2])
+			if (cpu_map__compare_aggr_cpu_id(s1, c->map[s2]))
 				break;
 		}
 		if (s2 == c->nr) {
@@ -153,9 +187,8 @@ int cpu_map__build_map(struct perf_cpu_map *cpus, struct perf_cpu_map **res,
 		}
 	}
 	/* ensure we process id in increasing order */
-	qsort(c->map, c->nr, sizeof(int), cmp_ids);
+	qsort(c->map, c->nr, sizeof(struct aggr_cpu_id), cmp_aggr_cpu_id);
 
-	refcount_set(&c->refcnt, 1);
 	*res = c;
 	return 0;
 }
@@ -167,37 +200,32 @@ int cpu_map__get_die_id(int cpu)
 	return ret ?: value;
 }
 
-int cpu_map__get_die(struct perf_cpu_map *map, int idx, void *data)
+struct aggr_cpu_id cpu_map__get_die(struct perf_cpu_map *map, int idx, void *data)
 {
-	int cpu, die_id, s;
+	int cpu, die;
+	struct aggr_cpu_id id = cpu_map__empty_aggr_cpu_id();
 
 	if (idx > map->nr)
-		return -1;
+		return id;
 
 	cpu = map->map[idx];
 
-	die_id = cpu_map__get_die_id(cpu);
+	die = cpu_map__get_die_id(cpu);
 	/* There is no die_id on legacy system. */
-	if (die_id == -1)
-		die_id = 0;
-
-	s = cpu_map__get_socket(map, idx, data);
-	if (s == -1)
-		return -1;
+	if (die == -1)
+		die = 0;
 
 	/*
-	 * Encode socket in bit range 15:8
-	 * die_id is relative to socket, and
-	 * we need a global id. So we combine
-	 * socket + die id
+	 * die_id is relative to socket, so start
+	 * with the socket ID and then add die to
+	 * make a unique ID.
 	 */
-	if (WARN_ONCE(die_id >> 8, "The die id number is too big.\n"))
-		return -1;
-
-	if (WARN_ONCE(s >> 8, "The socket id number is too big.\n"))
-		return -1;
+	id = cpu_map__get_socket(map, idx, data);
+	if (cpu_map__aggr_cpu_id_is_empty(id))
+		return id;
 
-	return (s << 8) | (die_id & 0xff);
+	id.die = die;
+	return id;
 }
 
 int cpu_map__get_core_id(int cpu)
@@ -211,59 +239,58 @@ int cpu_map__get_node_id(int cpu)
 	return cpu__get_node(cpu);
 }
 
-int cpu_map__get_core(struct perf_cpu_map *map, int idx, void *data)
+struct aggr_cpu_id cpu_map__get_core(struct perf_cpu_map *map, int idx, void *data)
 {
-	int cpu, s_die;
+	int cpu;
+	struct aggr_cpu_id id = cpu_map__empty_aggr_cpu_id();
 
 	if (idx > map->nr)
-		return -1;
+		return id;
 
 	cpu = map->map[idx];
 
 	cpu = cpu_map__get_core_id(cpu);
 
-	/* s_die is the combination of socket + die id */
-	s_die = cpu_map__get_die(map, idx, data);
-	if (s_die == -1)
-		return -1;
+	/* cpu_map__get_die returns a struct with socket and die set*/
+	id = cpu_map__get_die(map, idx, data);
+	if (cpu_map__aggr_cpu_id_is_empty(id))
+		return id;
 
 	/*
-	 * 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 + die id + core id
+	 * core_id is relative to socket and die, we need a global id.
+	 * So we combine the result from cpu_map__get_die with the core id
 	 */
-	if (WARN_ONCE(cpu >> 16, "The core id number is too big.\n"))
-		return -1;
-
-	return (s_die << 16) | (cpu & 0xffff);
+	id.core = cpu;
+	return id;
 }
 
-int cpu_map__get_node(struct perf_cpu_map *map, int idx, void *data __maybe_unused)
+struct aggr_cpu_id cpu_map__get_node(struct perf_cpu_map *map, int idx, void *data __maybe_unused)
 {
+	struct aggr_cpu_id id = cpu_map__empty_aggr_cpu_id();
+
 	if (idx < 0 || idx >= map->nr)
-		return -1;
+		return id;
 
-	return cpu_map__get_node_id(map->map[idx]);
+	id.node = cpu_map__get_node_id(map->map[idx]);
+	return id;
 }
 
-int cpu_map__build_socket_map(struct perf_cpu_map *cpus, struct perf_cpu_map **sockp)
+int cpu_map__build_socket_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **sockp)
 {
 	return cpu_map__build_map(cpus, sockp, cpu_map__get_socket, NULL);
 }
 
-int cpu_map__build_die_map(struct perf_cpu_map *cpus, struct perf_cpu_map **diep)
+int cpu_map__build_die_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **diep)
 {
 	return cpu_map__build_map(cpus, diep, cpu_map__get_die, NULL);
 }
 
-int cpu_map__build_core_map(struct perf_cpu_map *cpus, struct perf_cpu_map **corep)
+int cpu_map__build_core_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **corep)
 {
 	return cpu_map__build_map(cpus, corep, cpu_map__get_core, NULL);
 }
 
-int cpu_map__build_node_map(struct perf_cpu_map *cpus, struct perf_cpu_map **numap)
+int cpu_map__build_node_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **numap)
 {
 	return cpu_map__build_map(cpus, numap, cpu_map__get_node, NULL);
 }
@@ -586,3 +613,33 @@ const struct perf_cpu_map *cpu_map__online(void) /* thread unsafe */
 
 	return online;
 }
+
+bool cpu_map__compare_aggr_cpu_id(struct aggr_cpu_id a, struct aggr_cpu_id b)
+{
+	return a.thread == b.thread &&
+		a.node == b.node &&
+		a.socket == b.socket &&
+		a.die == b.die &&
+		a.core == b.core;
+}
+
+bool cpu_map__aggr_cpu_id_is_empty(struct aggr_cpu_id a)
+{
+	return a.thread == -1 &&
+		a.node == -1 &&
+		a.socket == -1 &&
+		a.die == -1 &&
+		a.core == -1;
+}
+
+struct aggr_cpu_id cpu_map__empty_aggr_cpu_id(void)
+{
+	struct aggr_cpu_id ret = {
+		.thread = -1,
+		.node = -1,
+		.socket = -1,
+		.die = -1,
+		.core = -1
+	};
+	return ret;
+}
diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h
index 3a442f021468..a27eeaf086e8 100644
--- a/tools/perf/util/cpumap.h
+++ b/tools/perf/util/cpumap.h
@@ -7,25 +7,41 @@
 #include <internal/cpumap.h>
 #include <perf/cpumap.h>
 
+struct aggr_cpu_id {
+	int thread;
+	int node;
+	int socket;
+	int die;
+	int core;
+};
+
+struct cpu_aggr_map {
+	refcount_t refcnt;
+	int nr;
+	struct aggr_cpu_id map[];
+};
+
 struct perf_record_cpu_map_data;
 
 struct perf_cpu_map *perf_cpu_map__empty_new(int nr);
+struct cpu_aggr_map *cpu_aggr_map__empty_new(int nr);
+
 struct perf_cpu_map *cpu_map__new_data(struct perf_record_cpu_map_data *data);
 size_t cpu_map__snprint(struct perf_cpu_map *map, char *buf, size_t size);
 size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size);
 size_t cpu_map__fprintf(struct perf_cpu_map *map, FILE *fp);
 int cpu_map__get_socket_id(int cpu);
-int cpu_map__get_socket(struct perf_cpu_map *map, int idx, void *data);
+struct aggr_cpu_id cpu_map__get_socket(struct perf_cpu_map *map, int idx, void *data);
 int cpu_map__get_die_id(int cpu);
-int cpu_map__get_die(struct perf_cpu_map *map, int idx, void *data);
+struct aggr_cpu_id cpu_map__get_die(struct perf_cpu_map *map, int idx, void *data);
 int cpu_map__get_core_id(int cpu);
-int cpu_map__get_core(struct perf_cpu_map *map, int idx, void *data);
+struct aggr_cpu_id cpu_map__get_core(struct perf_cpu_map *map, int idx, void *data);
 int cpu_map__get_node_id(int cpu);
-int cpu_map__get_node(struct perf_cpu_map *map, int idx, void *data);
-int cpu_map__build_socket_map(struct perf_cpu_map *cpus, struct perf_cpu_map **sockp);
-int cpu_map__build_die_map(struct perf_cpu_map *cpus, struct perf_cpu_map **diep);
-int cpu_map__build_core_map(struct perf_cpu_map *cpus, struct perf_cpu_map **corep);
-int cpu_map__build_node_map(struct perf_cpu_map *cpus, struct perf_cpu_map **nodep);
+struct aggr_cpu_id  cpu_map__get_node(struct perf_cpu_map *map, int idx, void *data);
+int cpu_map__build_socket_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **sockp);
+int cpu_map__build_die_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **diep);
+int cpu_map__build_core_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **corep);
+int cpu_map__build_node_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **nodep);
 const struct perf_cpu_map *cpu_map__online(void); /* thread unsafe */
 
 static inline int cpu_map__socket(struct perf_cpu_map *sock, int s)
@@ -35,21 +51,6 @@ static inline int cpu_map__socket(struct perf_cpu_map *sock, int s)
 	return sock->map[s];
 }
 
-static inline int cpu_map__id_to_socket(int id)
-{
-	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)
-{
-	return id & 0xffff;
-}
-
 int cpu__setup_cpunode_map(void);
 
 int cpu__max_node(void);
@@ -57,11 +58,15 @@ int cpu__max_cpu(void);
 int cpu__max_present_cpu(void);
 int cpu__get_node(int cpu);
 
-int cpu_map__build_map(struct perf_cpu_map *cpus, struct perf_cpu_map **res,
-		       int (*f)(struct perf_cpu_map *map, int cpu, void *data),
+int cpu_map__build_map(struct perf_cpu_map *cpus, struct cpu_aggr_map **res,
+		       struct aggr_cpu_id (*f)(struct perf_cpu_map *map, int cpu, void *data),
 		       void *data);
 
 int cpu_map__cpu(struct perf_cpu_map *cpus, int idx);
 bool cpu_map__has(struct perf_cpu_map *cpus, int cpu);
 
+bool cpu_map__compare_aggr_cpu_id(struct aggr_cpu_id a, struct aggr_cpu_id b);
+bool cpu_map__aggr_cpu_id_is_empty(struct aggr_cpu_id a);
+struct aggr_cpu_id cpu_map__empty_aggr_cpu_id(void);
+
 #endif /* __PERF_CPUMAP_H */
diff --git a/tools/perf/util/probe-file.c b/tools/perf/util/probe-file.c
index 064b63a6a3f3..bbecb449ea94 100644
--- a/tools/perf/util/probe-file.c
+++ b/tools/perf/util/probe-file.c
@@ -791,7 +791,7 @@ static char *synthesize_sdt_probe_command(struct sdt_note *note,
 					const char *sdtgrp)
 {
 	struct strbuf buf;
-	char *ret = NULL, **args;
+	char *ret = NULL;
 	int i, args_count, err;
 	unsigned long long ref_ctr_offset;
 
@@ -813,12 +813,19 @@ static char *synthesize_sdt_probe_command(struct sdt_note *note,
 		goto out;
 
 	if (note->args) {
-		args = argv_split(note->args, &args_count);
+		char **args = argv_split(note->args, &args_count);
+
+		if (args == NULL)
+			goto error;
 
 		for (i = 0; i < args_count; ++i) {
-			if (synthesize_sdt_probe_arg(&buf, i, args[i]) < 0)
+			if (synthesize_sdt_probe_arg(&buf, i, args[i]) < 0) {
+				argv_free(args);
 				goto error;
+			}
 		}
+
+		argv_free(args);
 	}
 
 out:
diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c
index fee7543843a8..583ae4f09c5d 100644
--- a/tools/perf/util/stat-display.c
+++ b/tools/perf/util/stat-display.c
@@ -68,15 +68,15 @@ static void print_cgroup(struct perf_stat_config *config, struct evsel *evsel)
 
 
 static void aggr_printout(struct perf_stat_config *config,
-			  struct evsel *evsel, int id, int nr)
+			  struct evsel *evsel, struct aggr_cpu_id id, int nr)
 {
 	switch (config->aggr_mode) {
 	case AGGR_CORE:
 		fprintf(config->output, "S%d-D%d-C%*d%s%*d%s",
-			cpu_map__id_to_socket(id),
-			cpu_map__id_to_die(id),
+			id.socket,
+			id.die,
 			config->csv_output ? 0 : -8,
-			cpu_map__id_to_cpu(id),
+			id.core,
 			config->csv_sep,
 			config->csv_output ? 0 : 4,
 			nr,
@@ -84,9 +84,9 @@ static void aggr_printout(struct perf_stat_config *config,
 		break;
 	case AGGR_DIE:
 		fprintf(config->output, "S%d-D%*d%s%*d%s",
-			cpu_map__id_to_socket(id << 16),
+			id.socket,
 			config->csv_output ? 0 : -8,
-			cpu_map__id_to_die(id << 16),
+			id.die,
 			config->csv_sep,
 			config->csv_output ? 0 : 4,
 			nr,
@@ -95,7 +95,7 @@ static void aggr_printout(struct perf_stat_config *config,
 	case AGGR_SOCKET:
 		fprintf(config->output, "S%*d%s%*d%s",
 			config->csv_output ? 0 : -5,
-			id,
+			id.socket,
 			config->csv_sep,
 			config->csv_output ? 0 : 4,
 			nr,
@@ -104,7 +104,7 @@ static void aggr_printout(struct perf_stat_config *config,
 	case AGGR_NODE:
 		fprintf(config->output, "N%*d%s%*d%s",
 			config->csv_output ? 0 : -5,
-			id,
+			id.node,
 			config->csv_sep,
 			config->csv_output ? 0 : 4,
 			nr,
@@ -113,23 +113,23 @@ static void aggr_printout(struct perf_stat_config *config,
 	case AGGR_NONE:
 		if (evsel->percore && !config->percore_show_thread) {
 			fprintf(config->output, "S%d-D%d-C%*d%s",
-				cpu_map__id_to_socket(id),
-				cpu_map__id_to_die(id),
+				id.socket,
+				id.die,
 				config->csv_output ? 0 : -3,
-				cpu_map__id_to_cpu(id), config->csv_sep);
-		} else if (id > -1) {
+				id.core, config->csv_sep);
+		} else if (id.core > -1) {
 			fprintf(config->output, "CPU%*d%s",
 				config->csv_output ? 0 : -7,
-				evsel__cpus(evsel)->map[id],
+				evsel__cpus(evsel)->map[id.core],
 				config->csv_sep);
 		}
 		break;
 	case AGGR_THREAD:
 		fprintf(config->output, "%*s-%*d%s",
 			config->csv_output ? 0 : 16,
-			perf_thread_map__comm(evsel->core.threads, id),
+			perf_thread_map__comm(evsel->core.threads, id.thread),
 			config->csv_output ? 0 : -8,
-			perf_thread_map__pid(evsel->core.threads, id),
+			perf_thread_map__pid(evsel->core.threads, id.thread),
 			config->csv_sep);
 		break;
 	case AGGR_GLOBAL:
@@ -144,7 +144,8 @@ struct outstate {
 	bool newline;
 	const char *prefix;
 	int  nfields;
-	int  id, nr;
+	int  nr;
+	struct aggr_cpu_id id;
 	struct evsel *evsel;
 };
 
@@ -319,13 +320,13 @@ static void print_metric_header(struct perf_stat_config *config,
 }
 
 static int first_shadow_cpu(struct perf_stat_config *config,
-			    struct evsel *evsel, int id)
+			    struct evsel *evsel, struct aggr_cpu_id id)
 {
 	struct evlist *evlist = evsel->evlist;
 	int i;
 
 	if (config->aggr_mode == AGGR_NONE)
-		return id;
+		return id.core;
 
 	if (!config->aggr_get_id)
 		return 0;
@@ -333,14 +334,17 @@ static int first_shadow_cpu(struct perf_stat_config *config,
 	for (i = 0; i < evsel__nr_cpus(evsel); i++) {
 		int cpu2 = evsel__cpus(evsel)->map[i];
 
-		if (config->aggr_get_id(config, evlist->core.cpus, cpu2) == id)
+		if (cpu_map__compare_aggr_cpu_id(
+					config->aggr_get_id(config, evlist->core.cpus, cpu2),
+					id)) {
 			return cpu2;
+		}
 	}
 	return 0;
 }
 
 static void abs_printout(struct perf_stat_config *config,
-			 int id, int nr, struct evsel *evsel, double avg)
+			 struct aggr_cpu_id id, int nr, struct evsel *evsel, double avg)
 {
 	FILE *output = config->output;
 	double sc =  evsel->scale;
@@ -393,7 +397,7 @@ static bool is_mixed_hw_group(struct evsel *counter)
 	return false;
 }
 
-static void printout(struct perf_stat_config *config, int id, int nr,
+static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int nr,
 		     struct evsel *counter, double uval,
 		     char *prefix, u64 run, u64 ena, double noise,
 		     struct runtime_stat *st)
@@ -496,7 +500,8 @@ static void printout(struct perf_stat_config *config, int id, int nr,
 static void aggr_update_shadow(struct perf_stat_config *config,
 			       struct evlist *evlist)
 {
-	int cpu, s2, id, s;
+	int cpu, s;
+	struct aggr_cpu_id s2, id;
 	u64 val;
 	struct evsel *counter;
 
@@ -506,7 +511,7 @@ static void aggr_update_shadow(struct perf_stat_config *config,
 			val = 0;
 			for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
 				s2 = config->aggr_get_id(config, evlist->core.cpus, cpu);
-				if (s2 != id)
+				if (!cpu_map__compare_aggr_cpu_id(s2, id))
 					continue;
 				val += perf_counts(counter->counts, cpu, 0)->val;
 			}
@@ -584,7 +589,7 @@ static bool collect_data(struct perf_stat_config *config, struct evsel *counter,
 
 struct aggr_data {
 	u64 ena, run, val;
-	int id;
+	struct aggr_cpu_id id;
 	int nr;
 	int cpu;
 };
@@ -593,13 +598,14 @@ static void aggr_cb(struct perf_stat_config *config,
 		    struct evsel *counter, void *data, bool first)
 {
 	struct aggr_data *ad = data;
-	int cpu, s2;
+	int cpu;
+	struct aggr_cpu_id s2;
 
 	for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
 		struct perf_counts_values *counts;
 
 		s2 = config->aggr_get_id(config, evsel__cpus(counter), cpu);
-		if (s2 != ad->id)
+		if (!cpu_map__compare_aggr_cpu_id(s2, ad->id))
 			continue;
 		if (first)
 			ad->nr++;
@@ -628,7 +634,8 @@ static void print_counter_aggrdata(struct perf_stat_config *config,
 	struct aggr_data ad;
 	FILE *output = config->output;
 	u64 ena, run, val;
-	int id, nr;
+	int nr;
+	struct aggr_cpu_id id;
 	double uval;
 
 	ad.id = id = config->aggr_map->map[s];
@@ -649,8 +656,12 @@ static void print_counter_aggrdata(struct perf_stat_config *config,
 		fprintf(output, "%s", prefix);
 
 	uval = val * counter->scale;
-	printout(config, cpu != -1 ? cpu : id, nr, counter, uval, prefix,
-		 run, ena, 1.0, &rt_stat);
+	if (cpu != -1) {
+		id = cpu_map__empty_aggr_cpu_id();
+		id.core = cpu;
+	}
+	printout(config, id, nr, counter, uval,
+		 prefix, run, ena, 1.0, &rt_stat);
 	if (!metric_only)
 		fputc('\n', output);
 }
@@ -728,7 +739,8 @@ static struct perf_aggr_thread_value *sort_aggr_thread(
 			continue;
 
 		buf[i].counter = counter;
-		buf[i].id = thread;
+		buf[i].id = cpu_map__empty_aggr_cpu_id();
+		buf[i].id.thread = thread;
 		buf[i].uval = uval;
 		buf[i].val = val;
 		buf[i].run = run;
@@ -751,7 +763,8 @@ static void print_aggr_thread(struct perf_stat_config *config,
 	FILE *output = config->output;
 	int nthreads = perf_thread_map__nr(counter->core.threads);
 	int ncpus = perf_cpu_map__nr(counter->core.cpus);
-	int thread, sorted_threads, id;
+	int thread, sorted_threads;
+	struct aggr_cpu_id id;
 	struct perf_aggr_thread_value *buf;
 
 	buf = sort_aggr_thread(counter, nthreads, ncpus, &sorted_threads, _target);
@@ -768,7 +781,7 @@ static void print_aggr_thread(struct perf_stat_config *config,
 		if (config->stats)
 			printout(config, id, 0, buf[thread].counter, buf[thread].uval,
 				 prefix, buf[thread].run, buf[thread].ena, 1.0,
-				 &config->stats[id]);
+				 &config->stats[id.thread]);
 		else
 			printout(config, id, 0, buf[thread].counter, buf[thread].uval,
 				 prefix, buf[thread].run, buf[thread].ena, 1.0,
@@ -814,8 +827,8 @@ static void print_counter_aggr(struct perf_stat_config *config,
 		fprintf(output, "%s", prefix);
 
 	uval = cd.avg * counter->scale;
-	printout(config, -1, 0, counter, uval, prefix, cd.avg_running, cd.avg_enabled,
-		 cd.avg, &rt_stat);
+	printout(config, cpu_map__empty_aggr_cpu_id(), 0, counter, uval, prefix, cd.avg_running,
+		 cd.avg_enabled, cd.avg, &rt_stat);
 	if (!metric_only)
 		fprintf(output, "\n");
 }
@@ -842,6 +855,7 @@ static void print_counter(struct perf_stat_config *config,
 	u64 ena, run, val;
 	double uval;
 	int cpu;
+	struct aggr_cpu_id id;
 
 	for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
 		struct aggr_data ad = { .cpu = cpu };
@@ -856,8 +870,10 @@ static void print_counter(struct perf_stat_config *config,
 			fprintf(output, "%s", prefix);
 
 		uval = val * counter->scale;
-		printout(config, cpu, 0, counter, uval, prefix, run, ena, 1.0,
-			 &rt_stat);
+		id = cpu_map__empty_aggr_cpu_id();
+		id.core = cpu;
+		printout(config, id, 0, counter, uval, prefix,
+			 run, ena, 1.0, &rt_stat);
 
 		fputc('\n', output);
 	}
@@ -872,6 +888,7 @@ static void print_no_aggr_metric(struct perf_stat_config *config,
 	struct evsel *counter;
 	u64 ena, run, val;
 	double uval;
+	struct aggr_cpu_id id;
 
 	nrcpus = evlist->core.cpus->nr;
 	for (cpu = 0; cpu < nrcpus; cpu++) {
@@ -880,8 +897,10 @@ static void print_no_aggr_metric(struct perf_stat_config *config,
 		if (prefix)
 			fputs(prefix, config->output);
 		evlist__for_each_entry(evlist, counter) {
+			id = cpu_map__empty_aggr_cpu_id();
+			id.core = cpu;
 			if (first) {
-				aggr_printout(config, counter, cpu, 0);
+				aggr_printout(config, counter, id, 0);
 				first = false;
 			}
 			val = perf_counts(counter->counts, cpu, 0)->val;
@@ -889,8 +908,8 @@ static void print_no_aggr_metric(struct perf_stat_config *config,
 			run = perf_counts(counter->counts, cpu, 0)->run;
 
 			uval = val * counter->scale;
-			printout(config, cpu, 0, counter, uval, prefix, run, ena, 1.0,
-				 &rt_stat);
+			printout(config, id, 0, counter, uval, prefix,
+				 run, ena, 1.0, &rt_stat);
 		}
 		fputc('\n', config->output);
 	}
@@ -1140,14 +1159,15 @@ static void print_footer(struct perf_stat_config *config)
 static void print_percore_thread(struct perf_stat_config *config,
 				 struct evsel *counter, char *prefix)
 {
-	int s, s2, id;
+	int s;
+	struct aggr_cpu_id s2, id;
 	bool first = true;
 
 	for (int i = 0; i < evsel__nr_cpus(counter); i++) {
 		s2 = config->aggr_get_id(config, evsel__cpus(counter), i);
 		for (s = 0; s < config->aggr_map->nr; s++) {
 			id = config->aggr_map->map[s];
-			if (s2 == id)
+			if (cpu_map__compare_aggr_cpu_id(s2, id))
 				break;
 		}
 
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
index 1e125e39ff84..8ce1479c98f0 100644
--- a/tools/perf/util/stat.c
+++ b/tools/perf/util/stat.c
@@ -313,7 +313,7 @@ static int check_per_pkg(struct evsel *counter,
 	if (!(vals->run && vals->ena))
 		return 0;
 
-	s = cpu_map__get_socket(cpus, cpu, NULL);
+	s = cpu_map__get_socket(cpus, cpu, NULL).socket;
 	if (s < 0)
 		return -1;
 
diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h
index 9979b4b100f2..b5369730b4a2 100644
--- a/tools/perf/util/stat.h
+++ b/tools/perf/util/stat.h
@@ -6,6 +6,7 @@
 #include <stdio.h>
 #include <sys/types.h>
 #include <sys/resource.h>
+#include "cpumap.h"
 #include "rblist.h"
 
 struct perf_cpu_map;
@@ -99,7 +100,7 @@ struct runtime_stat {
 	struct rblist value_list;
 };
 
-typedef int (*aggr_get_id_t)(struct perf_stat_config *config,
+typedef struct aggr_cpu_id (*aggr_get_id_t)(struct perf_stat_config *config,
 			     struct perf_cpu_map *m, int cpu);
 
 struct perf_stat_config {
@@ -138,9 +139,9 @@ struct perf_stat_config {
 	const char		*csv_sep;
 	struct stats		*walltime_nsecs_stats;
 	struct rusage		 ru_data;
-	struct perf_cpu_map		*aggr_map;
+	struct cpu_aggr_map	*aggr_map;
 	aggr_get_id_t		 aggr_get_id;
-	struct perf_cpu_map		*cpus_aggr_map;
+	struct cpu_aggr_map	*cpus_aggr_map;
 	u64			*walltime_run;
 	struct rblist		 metric_events;
 	int			 ctl_fd;
@@ -170,7 +171,7 @@ struct evlist;
 
 struct perf_aggr_thread_value {
 	struct evsel *counter;
-	int id;
+	struct aggr_cpu_id id;
 	double uval;
 	u64 val;
 	u64 run;