perf sample: Add evsel to struct perf_sample

Add the evsel from evsel__parse_sample into the struct
perf_sample. Sometimes we want to alter the evsel associated with a
sample, such as with off-cpu bpf-output events. In general the evsel
and perf_sample are passed as a pair, but this makes an altered evsel
something of a chore to keep checking for and setting up. Later
patches will remove passing an evsel with the perf_sample and switch
to just using the perf_sample's value.

Signed-off-by: Ian Rogers <irogers@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>

tools/perf/builtin-inject.c

@@ -133,7 +133,7 @@ struct perf_inject {
 	struct perf_file_section secs[HEADER_FEAT_BITS];
 	struct guest_session	guest_session;
 	struct strlist		*known_build_ids;
-	const struct evsel	*mmap_evsel;
+	struct evsel		*mmap_evsel;
 	struct ip_callchain	*raw_callchain;
 };
 
@@ -519,7 +519,7 @@ static struct dso *findnew_dso(int pid, int tid, const char *filename,
  * processing mmap events. If not stashed, search the evlist for the first mmap
  * gathering event.
  */
-static const struct evsel *inject__mmap_evsel(struct perf_inject *inject)
+static struct evsel *inject__mmap_evsel(struct perf_inject *inject)
 {
 	struct evsel *pos;
 
@@ -1023,7 +1023,6 @@ int perf_event__inject_buildid(const struct perf_tool *tool, union perf_event *e
 
 	sample__for_each_callchain_node(thread, evsel, sample, PERF_MAX_STACK_DEPTH,
 					/*symbols=*/false, mark_dso_hit_callback, &args);
-
 	thread__put(thread);
 repipe:
 	perf_event__repipe(tool, event, sample, machine);
@@ -1432,6 +1431,7 @@ static int synthesize_build_id(struct perf_inject *inject, struct dso *dso, pid_
 {
 	struct machine *machine = perf_session__findnew_machine(inject->session, machine_pid);
 	struct perf_sample synth_sample = {
+		.evsel	   = inject__mmap_evsel(inject),
 		.pid	   = -1,
 		.tid	   = -1,
 		.time	   = -1,

tools/perf/builtin-script.c

@@ -2910,8 +2910,12 @@ static int print_event_with_time(const struct perf_tool *tool,
 		thread = machine__findnew_thread(machine, pid, tid);
 
 	if (evsel) {
+		struct evsel *saved_evsel = sample->evsel;
+
+		sample->evsel = evsel;
 		perf_sample__fprintf_start(script, sample, thread, evsel,
 					   event->header.type, stdout);
+		sample->evsel = saved_evsel;
 	}
 
 	perf_event__fprintf(event, machine, stdout);

tools/perf/tests/hists_cumulate.c

@@ -81,7 +81,7 @@ static int add_hist_entries(struct hists *hists, struct machine *machine)
 {
 	struct addr_location al;
 	struct evsel *evsel = hists_to_evsel(hists);
-	struct perf_sample sample = { .period = 1000, };
+	struct perf_sample sample = { .evsel = evsel, .period = 1000, };
 	size_t i;
 
 	addr_location__init(&al);

tools/perf/tests/hists_filter.c

@@ -70,6 +70,7 @@ static int add_hist_entries(struct evlist *evlist,
 			};
 			struct hists *hists = evsel__hists(evsel);
 
+			sample.evsel = evsel;
 			/* make sure it has no filter at first */
 			hists->thread_filter = NULL;
 			hists->dso_filter = NULL;

tools/perf/tests/hists_output.c

@@ -51,7 +51,7 @@ static int add_hist_entries(struct hists *hists, struct machine *machine)
 {
 	struct addr_location al;
 	struct evsel *evsel = hists_to_evsel(hists);
-	struct perf_sample sample = { .period = 100, };
+	struct perf_sample sample = { .evsel = evsel, .period = 100, };
 	size_t i;
 
 	addr_location__init(&al);

tools/perf/util/evsel.c

@@ -3226,6 +3226,7 @@ int evsel__parse_sample(struct evsel *evsel, union perf_event *event,
 	union u64_swap u;
 
 	perf_sample__init(data, /*all=*/true);
+	data->evsel = evsel;
 	data->cpu = data->pid = data->tid = -1;
 	data->stream_id = data->id = data->time = -1ULL;
 	data->period = evsel->core.attr.sample_period;

tools/perf/util/sample.c

@@ -19,6 +19,7 @@ void perf_sample__init(struct perf_sample *sample, bool all)
 	if (all) {
 		memset(sample, 0, sizeof(*sample));
 	} else {
+		sample->evsel = NULL;
 		sample->user_regs = NULL;
 		sample->intr_regs = NULL;
 		sample->merged_callchain = false;

tools/perf/util/sample.h

@@ -5,6 +5,7 @@
 #include <linux/perf_event.h>
 #include <linux/types.h>
 
+struct evsel;
 struct machine;
 struct thread;
 
@@ -102,6 +103,8 @@ struct simd_flags {
  * and clean up these values.
  */
 struct perf_sample {
+	/** @evsel: Backward reference to the evsel used when constructing the sample. */
+	struct evsel *evsel;
 	/** @ip: The sample event PERF_SAMPLE_IP value. */
 	u64 ip;
 	/** @pid: The sample event PERF_SAMPLE_TID pid value. */

tools/perf/util/session.c

@@ -1264,8 +1264,9 @@ static int deliver_sample_value(struct evlist *evlist,
 				bool per_thread)
 {
 	struct perf_sample_id *sid = evlist__id2sid(evlist, v->id);
-	struct evsel *evsel;
+	struct evsel *saved_evsel = sample->evsel;
 	u64 *storage = NULL;
+	int ret;
 
 	if (sid) {
 		storage = perf_sample_id__get_period_storage(sid, sample->tid, per_thread);
@@ -1289,8 +1290,10 @@ static int deliver_sample_value(struct evlist *evlist,
 	if (!sample->period)
 		return 0;
 
-	evsel = container_of(sid->evsel, struct evsel, core);
-	return tool->sample(tool, event, sample, evsel, machine);
+	sample->evsel = container_of(sid->evsel, struct evsel, core);
+	ret = tool->sample(tool, event, sample, sample->evsel, machine);
+	sample->evsel = saved_evsel;
+	return ret;
 }
 
 static int deliver_sample_group(struct evlist *evlist,
@@ -1362,13 +1365,16 @@ static int evlist__deliver_deferred_callchain(struct evlist *evlist,
 					      struct machine *machine)
 {
 	struct deferred_event *de, *tmp;
-	struct evsel *evsel;
 	int ret = 0;
 
 	if (!tool->merge_deferred_callchains) {
-		evsel = evlist__id2evsel(evlist, sample->id);
-		return tool->callchain_deferred(tool, event, sample,
-						evsel, machine);
+		struct evsel *saved_evsel = sample->evsel;
+
+		sample->evsel = evlist__id2evsel(evlist, sample->id);
+		ret = tool->callchain_deferred(tool, event, sample,
+					       sample->evsel, machine);
+		sample->evsel = saved_evsel;
+		return ret;
 	}
 
 	list_for_each_entry_safe(de, tmp, &evlist->deferred_samples, list) {
@@ -1392,9 +1398,9 @@ static int evlist__deliver_deferred_callchain(struct evlist *evlist,
 		else
 			orig_sample.deferred_callchain = false;
 
-		evsel = evlist__id2evsel(evlist, orig_sample.id);
+		orig_sample.evsel = evlist__id2evsel(evlist, orig_sample.id);
 		ret = evlist__deliver_sample(evlist, tool, de->event,
-					     &orig_sample, evsel, machine);
+					     &orig_sample, orig_sample.evsel, machine);
 
 		perf_sample__exit(&orig_sample);
 		list_del(&de->list);
@@ -1417,7 +1423,6 @@ static int session__flush_deferred_samples(struct perf_session *session,
 	struct evlist *evlist = session->evlist;
 	struct machine *machine = &session->machines.host;
 	struct deferred_event *de, *tmp;
-	struct evsel *evsel;
 	int ret = 0;
 
 	list_for_each_entry_safe(de, tmp, &evlist->deferred_samples, list) {
@@ -1431,9 +1436,9 @@ static int session__flush_deferred_samples(struct perf_session *session,
 			break;
 		}
 
-		evsel = evlist__id2evsel(evlist, sample.id);
+		sample.evsel = evlist__id2evsel(evlist, sample.id);
 		ret = evlist__deliver_sample(evlist, tool, de->event,
-					     &sample, evsel, machine);
+					     &sample, sample.evsel, machine);
 
 		perf_sample__exit(&sample);
 		list_del(&de->list);
@@ -1458,8 +1463,12 @@ static int machines__deliver_event(struct machines *machines,
 
 	dump_event(evlist, event, file_offset, sample, file_path);
 
-	evsel = evlist__id2evsel(evlist, sample->id);
+	if (!sample->evsel)
+		sample->evsel = evlist__id2evsel(evlist, sample->id);
+	else
+		assert(sample->evsel == evlist__id2evsel(evlist, sample->id));
 
+	evsel = sample->evsel;
 	machine = machines__find_for_cpumode(machines, event, sample);
 
 	switch (event->header.type) {