perf parse-events: Fix missing slots for Intel topdown metric events

Topdown metric events require grouping with a slots event. In perf
metrics this is currently achieved by metrics adding an unnecessary
"0 * tma_info_thread_slots". New TMA metrics trigger optimizations of
the metric expression that removes the event and breaks the metric due
to the missing but required event. Add a pass immediately before
sorting and fixing parsed events, that insert a slots event if one is
missing. Update test expectations to match this.

Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250719030517.1990983-15-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>

Author: captain5050

tools/perf/arch/x86/util/evlist.c

@@ -81,3 +81,27 @@ int arch_evlist__cmp(const struct evsel *lhs, const struct evsel *rhs)
 	/* Default ordering by insertion index. */
 	return lhs->core.idx - rhs->core.idx;
 }
+
+int arch_evlist__add_required_events(struct list_head *list)
+{
+	struct evsel *pos, *metric_event = NULL;
+	int idx = 0;
+
+	if (!topdown_sys_has_perf_metrics())
+		return 0;
+
+	list_for_each_entry(pos, list, core.node) {
+		if (arch_is_topdown_slots(pos)) {
+			/* Slots event already present, nothing to do. */
+			return 0;
+		}
+		if (metric_event == NULL && arch_is_topdown_metrics(pos))
+			metric_event = pos;
+		idx++;
+	}
+	if (metric_event == NULL) {
+		/* No topdown metric events, nothing to do. */
+		return 0;
+	}
+	return topdown_insert_slots_event(list, idx + 1, metric_event);
+}

tools/perf/arch/x86/util/topdown.c

@@ -77,3 +77,31 @@ bool arch_topdown_sample_read(struct evsel *leader)
 
 	return false;
 }
+
+/*
+ * Make a copy of the topdown metric event metric_event with the given index but
+ * change its configuration to be a topdown slots event. Copying from
+ * metric_event ensures modifiers are the same.
+ */
+int topdown_insert_slots_event(struct list_head *list, int idx, struct evsel *metric_event)
+{
+	struct evsel *evsel = evsel__new_idx(&metric_event->core.attr, idx);
+
+	if (!evsel)
+		return -ENOMEM;
+
+	evsel->core.attr.config = TOPDOWN_SLOTS;
+	evsel->core.cpus = perf_cpu_map__get(metric_event->core.cpus);
+	evsel->core.pmu_cpus = perf_cpu_map__get(metric_event->core.pmu_cpus);
+	evsel->core.is_pmu_core = true;
+	evsel->pmu = metric_event->pmu;
+	evsel->name = strdup("slots");
+	evsel->precise_max = metric_event->precise_max;
+	evsel->sample_read = metric_event->sample_read;
+	evsel->weak_group = metric_event->weak_group;
+	evsel->bpf_counter = metric_event->bpf_counter;
+	evsel->retire_lat = metric_event->retire_lat;
+	evsel__set_leader(evsel, evsel__leader(metric_event));
+	list_add_tail(&evsel->core.node, list);
+	return 0;
+}

tools/perf/arch/x86/util/topdown.h

@@ -5,9 +5,11 @@
 #include <stdbool.h>
 
 struct evsel;
+struct list_head;
 
 bool topdown_sys_has_perf_metrics(void);
 bool arch_is_topdown_slots(const struct evsel *evsel);
 bool arch_is_topdown_metrics(const struct evsel *evsel);
+int topdown_insert_slots_event(struct list_head *list, int idx, struct evsel *metric_event);
 
 #endif

tools/perf/tests/parse-events.c

@@ -719,20 +719,20 @@ static int test__checkevent_pmu_partial_time_callgraph(struct evlist *evlist)
 
 static int test__checkevent_pmu_events(struct evlist *evlist)
 {
-	struct evsel *evsel = evlist__first(evlist);
+	struct evsel *evsel;
 
-	TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->core.nr_entries);
-	TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->core.attr.type ||
-				      strcmp(evsel->pmu->name, "cpu"));
-	TEST_ASSERT_VAL("wrong exclude_user",
-			!evsel->core.attr.exclude_user);
-	TEST_ASSERT_VAL("wrong exclude_kernel",
-			evsel->core.attr.exclude_kernel);
-	TEST_ASSERT_VAL("wrong exclude_hv", evsel->core.attr.exclude_hv);
-	TEST_ASSERT_VAL("wrong precise_ip", !evsel->core.attr.precise_ip);
-	TEST_ASSERT_VAL("wrong pinned", !evsel->core.attr.pinned);
-	TEST_ASSERT_VAL("wrong exclusive", !evsel->core.attr.exclusive);
+	TEST_ASSERT_VAL("wrong number of entries", 1 <= evlist->core.nr_entries);
 
+	evlist__for_each_entry(evlist, evsel) {
+		TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->core.attr.type ||
+					      strcmp(evsel->pmu->name, "cpu"));
+		TEST_ASSERT_VAL("wrong exclude_user", !evsel->core.attr.exclude_user);
+		TEST_ASSERT_VAL("wrong exclude_kernel", evsel->core.attr.exclude_kernel);
+		TEST_ASSERT_VAL("wrong exclude_hv", evsel->core.attr.exclude_hv);
+		TEST_ASSERT_VAL("wrong precise_ip", !evsel->core.attr.precise_ip);
+		TEST_ASSERT_VAL("wrong pinned", !evsel->core.attr.pinned);
+		TEST_ASSERT_VAL("wrong exclusive", !evsel->core.attr.exclusive);
+	}
 	return TEST_OK;
 }
 

tools/perf/util/evlist.h

@@ -111,6 +111,7 @@ void evlist__add(struct evlist *evlist, struct evsel *entry);
 void evlist__remove(struct evlist *evlist, struct evsel *evsel);
 
 int arch_evlist__cmp(const struct evsel *lhs, const struct evsel *rhs);
+int arch_evlist__add_required_events(struct list_head *list);
 
 int evlist__add_dummy(struct evlist *evlist);
 struct evsel *evlist__add_aux_dummy(struct evlist *evlist, bool system_wide);

tools/perf/util/parse-events.c

@@ -2190,6 +2190,11 @@ static int evlist__cmp(void *_fg_idx, const struct list_head *l, const struct li
 	return arch_evlist__cmp(lhs, rhs);
 }
 
+int __weak arch_evlist__add_required_events(struct list_head *list __always_unused)
+{
+	return 0;
+}
+
 static int parse_events__sort_events_and_fix_groups(struct list_head *list)
 {
 	int idx = 0, force_grouped_idx = -1;
@@ -2201,6 +2206,11 @@ static int parse_events__sort_events_and_fix_groups(struct list_head *list)
 	struct evsel *force_grouped_leader = NULL;
 	bool last_event_was_forced_leader = false;
 
+	/* On x86 topdown metrics events require a slots event. */
+	ret = arch_evlist__add_required_events(list);
+	if (ret)
+		return ret;
+
 	/*
 	 * Compute index to insert ungrouped events at. Place them where the
 	 * first ungrouped event appears.