perf pmus: Allow just core PMU scanning

Scanning all PMUs is expensive as all PMUs sysfs entries are loaded,
benchmarking shows more than 4x the cost:

```
$ perf bench internals pmu-scan -i 1000
Computing performance of sysfs PMU event scan for 1000 times
  Average core PMU scanning took: 989.231 usec (+- 1.535 usec)
  Average PMU scanning took: 4309.425 usec (+- 74.322 usec)
```

Add new perf_pmus__scan_core routine that scans just core
PMUs. Replace perf_pmus__scan calls with perf_pmus__scan_core when
non-core PMUs are being ignored.

Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ali Saidi <alisaidi@amazon.com>
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Cc: Dmitrii Dolgov <9erthalion6@gmail.com>
Cc: Huacai Chen <chenhuacai@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@arm.com>
Cc: Jing Zhang <renyu.zj@linux.alibaba.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Garry <john.g.garry@oracle.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kang Minchul <tegongkang@gmail.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Ming Wang <wangming01@loongson.cn>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@amd.com>
Cc: Rob Herring <robh@kernel.org>
Cc: Sandipan Das <sandipan.das@amd.com>
Cc: Sean Christopherson <seanjc@google.com>
Cc: Suzuki Poulouse <suzuki.poulose@arm.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Xing Zhengjun <zhengjun.xing@linux.intel.com>
Cc: coresight@lists.linaro.org
Cc: linux-arm-kernel@lists.infradead.org
Link: https://lore.kernel.org/r/20230527072210.2900565-30-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

9 files changed, 40 insertions, 58 deletions

diff --git a/tools/perf/util/cputopo.c b/tools/perf/util/cputopo.c
index 4578c26747e1..729142ec9a9a 100644
--- a/tools/perf/util/cputopo.c
+++ b/tools/perf/util/cputopo.c
@@ -477,10 +477,9 @@ struct hybrid_topology *hybrid_topology__new(void)
 	if (!perf_pmus__has_hybrid())
 		return NULL;
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		if (pmu->is_core)
-			nr++;
-	}
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL)
+		nr++;
+
 	if (nr == 0)
 		return NULL;
 
@@ -489,10 +488,7 @@ struct hybrid_topology *hybrid_topology__new(void)
 		return NULL;
 
 	tp->nr = nr;
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		if (!pmu->is_core)
-			continue;
-
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 		if (load_hybrid_node(&tp->nodes[i], pmu)) {
 			hybrid_topology__delete(tp);
 			return NULL;
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index e6d8ecd7a08e..2dde3ca20de5 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -1607,10 +1607,7 @@ static int write_pmu_caps(struct feat_fd *ff,
 	 */
 	if (perf_pmus__has_hybrid()) {
 		pmu = NULL;
-		while ((pmu = perf_pmus__scan(pmu))) {
-			if (!pmu->is_core)
-				continue;
-
+		while ((pmu = perf_pmus__scan_core(pmu))) {
 			ret = __write_pmu_caps(ff, pmu, true);
 			if (ret < 0)
 				return ret;
diff --git a/tools/perf/util/mem-events.c b/tools/perf/util/mem-events.c
index 08ac3ea2e366..c5596230a308 100644
--- a/tools/perf/util/mem-events.c
+++ b/tools/perf/util/mem-events.c
@@ -136,10 +136,7 @@ int perf_mem_events__init(void)
 		} else {
 			struct perf_pmu *pmu = NULL;
 
-			while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-				if (!pmu->is_core)
-					continue;
-
+			while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 				scnprintf(sysfs_name, sizeof(sysfs_name),
 					  e->sysfs_name, pmu->name);
 				e->supported |= perf_mem_event__supported(mnt, sysfs_name);
@@ -176,10 +173,7 @@ static void perf_mem_events__print_unsupport_hybrid(struct perf_mem_event *e,
 	char sysfs_name[100];
 	struct perf_pmu *pmu = NULL;
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		if (!pmu->is_core)
-			continue;
-
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 		scnprintf(sysfs_name, sizeof(sysfs_name), e->sysfs_name,
 			  pmu->name);
 		if (!perf_mem_event__supported(mnt, sysfs_name)) {
@@ -217,9 +211,7 @@ int perf_mem_events__record_args(const char **rec_argv, int *argv_nr,
 				return -1;
 			}
 
-			while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-				if (!pmu->is_core)
-					continue;
+			while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 				rec_argv[i++] = "-e";
 				s = perf_mem_events__name(j, pmu->name);
 				if (s) {
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index be544f948be2..e0c3f2037477 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -453,15 +453,12 @@ int parse_events_add_cache(struct list_head *list, int *idx, const char *name,
 	const char *config_name = get_config_name(head_config);
 	const char *metric_id = get_config_metric_id(head_config);
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
+	/* Legacy cache events are only supported by core PMUs. */
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 		LIST_HEAD(config_terms);
 		struct perf_event_attr attr;
 		int ret;
 
-		/* Skip unsupported PMUs. */
-		if (!perf_pmu__supports_legacy_cache(pmu))
-			continue;
-
 		if (parse_events__filter_pmu(parse_state, pmu))
 			continue;
 
@@ -1481,12 +1478,10 @@ int parse_events_add_numeric(struct parse_events_state *parse_state,
 		return __parse_events_add_numeric(parse_state, list, /*pmu=*/NULL,
 						  type, config, head_config);
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
+	/* Wildcards on numeric values are only supported by core PMUs. */
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 		int ret;
 
-		if (!perf_pmu__supports_wildcard_numeric(pmu))
-			continue;
-
 		if (parse_events__filter_pmu(parse_state, pmu))
 			continue;
 
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 05056305fb58..7102084dd3aa 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -1427,21 +1427,11 @@ bool perf_pmu__supports_legacy_cache(const struct perf_pmu *pmu)
 	return pmu->is_core;
 }
 
-bool perf_pmu__supports_wildcard_numeric(const struct perf_pmu *pmu)
-{
-	return pmu->is_core;
-}
-
 bool perf_pmu__auto_merge_stats(const struct perf_pmu *pmu)
 {
 	return !is_pmu_hybrid(pmu->name);
 }
 
-bool perf_pmu__is_mem_pmu(const struct perf_pmu *pmu)
-{
-	return pmu->is_core;
-}
-
 bool perf_pmu__have_event(const struct perf_pmu *pmu, const char *name)
 {
 	struct perf_pmu_alias *alias;
diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h
index f1f3e8a2e00e..02fec0a7d4c8 100644
--- a/tools/perf/util/pmu.h
+++ b/tools/perf/util/pmu.h
@@ -223,9 +223,7 @@ void perf_pmu__del_formats(struct list_head *formats);
 bool is_pmu_core(const char *name);
 bool is_pmu_hybrid(const char *name);
 bool perf_pmu__supports_legacy_cache(const struct perf_pmu *pmu);
-bool perf_pmu__supports_wildcard_numeric(const struct perf_pmu *pmu);
 bool perf_pmu__auto_merge_stats(const struct perf_pmu *pmu);
-bool perf_pmu__is_mem_pmu(const struct perf_pmu *pmu);
 bool perf_pmu__have_event(const struct perf_pmu *pmu, const char *name);
 
 FILE *perf_pmu__open_file(struct perf_pmu *pmu, const char *name);
diff --git a/tools/perf/util/pmus.c b/tools/perf/util/pmus.c
index 4ef4fecd335f..de7fc36519c9 100644
--- a/tools/perf/util/pmus.c
+++ b/tools/perf/util/pmus.c
@@ -87,7 +87,7 @@ static struct perf_pmu *perf_pmu__find2(int dirfd, const char *name)
 }
 
 /* Add all pmus in sysfs to pmu list: */
-static void pmu_read_sysfs(void)
+static void pmu_read_sysfs(bool core_only)
 {
 	int fd;
 	DIR *dir;
@@ -104,6 +104,8 @@ static void pmu_read_sysfs(void)
 	while ((dent = readdir(dir))) {
 		if (!strcmp(dent->d_name, ".") || !strcmp(dent->d_name, ".."))
 			continue;
+		if (core_only && !is_pmu_core(dent->d_name))
+			continue;
 		/* add to static LIST_HEAD(core_pmus) or LIST_HEAD(other_pmus): */
 		perf_pmu__find2(fd, dent->d_name);
 	}
@@ -135,7 +137,7 @@ struct perf_pmu *perf_pmus__scan(struct perf_pmu *pmu)
 	bool use_core_pmus = !pmu || pmu->is_core;
 
 	if (!pmu) {
-		pmu_read_sysfs();
+		pmu_read_sysfs(/*core_only=*/false);
 		pmu = list_prepare_entry(pmu, &core_pmus, list);
 	}
 	if (use_core_pmus) {
@@ -150,6 +152,18 @@ struct perf_pmu *perf_pmus__scan(struct perf_pmu *pmu)
 	return NULL;
 }
 
+struct perf_pmu *perf_pmus__scan_core(struct perf_pmu *pmu)
+{
+	if (!pmu) {
+		pmu_read_sysfs(/*core_only=*/true);
+		pmu = list_prepare_entry(pmu, &core_pmus, list);
+	}
+	list_for_each_entry_continue(pmu, &core_pmus, list)
+		return pmu;
+
+	return NULL;
+}
+
 const struct perf_pmu *perf_pmus__pmu_for_pmu_filter(const char *str)
 {
 	struct perf_pmu *pmu = NULL;
@@ -176,10 +190,10 @@ int perf_pmus__num_mem_pmus(void)
 	struct perf_pmu *pmu = NULL;
 	int count = 0;
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		if (perf_pmu__is_mem_pmu(pmu))
-			count++;
-	}
+	/* All core PMUs are for mem events. */
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL)
+		count++;
+
 	return count;
 }
 
@@ -421,8 +435,8 @@ bool perf_pmus__has_hybrid(void)
 	if (!hybrid_scanned) {
 		struct perf_pmu *pmu = NULL;
 
-		while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-			if (pmu->is_core && is_pmu_hybrid(pmu->name)) {
+		while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
+			if (is_pmu_hybrid(pmu->name)) {
 				has_hybrid = true;
 				break;
 			}
diff --git a/tools/perf/util/pmus.h b/tools/perf/util/pmus.h
index 2a771d9f8da7..9de0222ed52b 100644
--- a/tools/perf/util/pmus.h
+++ b/tools/perf/util/pmus.h
@@ -11,6 +11,7 @@ struct perf_pmu *perf_pmus__find(const char *name);
 struct perf_pmu *perf_pmus__find_by_type(unsigned int type);
 
 struct perf_pmu *perf_pmus__scan(struct perf_pmu *pmu);
+struct perf_pmu *perf_pmus__scan_core(struct perf_pmu *pmu);
 
 const struct perf_pmu *perf_pmus__pmu_for_pmu_filter(const char *str);
 
diff --git a/tools/perf/util/print-events.c b/tools/perf/util/print-events.c
index 9cee7bb7a561..7a5f87392720 100644
--- a/tools/perf/util/print-events.c
+++ b/tools/perf/util/print-events.c
@@ -272,12 +272,11 @@ int print_hwcache_events(const struct print_callbacks *print_cb, void *print_sta
 	struct perf_pmu *pmu = NULL;
 	const char *event_type_descriptor = event_type_descriptors[PERF_TYPE_HW_CACHE];
 
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		/*
-		 * Skip uncore PMUs for performance. PERF_TYPE_HW_CACHE type
-		 * attributes can accept software PMUs in the extended type, so
-		 * also skip.
-		 */
+	/*
+	 * Only print core PMUs, skipping uncore for performance and
+	 * PERF_TYPE_SOFTWARE that can succeed in opening legacy cache evenst.
+	 */
+	while ((pmu = perf_pmus__scan_core(pmu)) != NULL) {
 		if (pmu->is_uncore || pmu->type == PERF_TYPE_SOFTWARE)
 			continue;