perf evsel: Fix missing exclude_{host,guest} setting

The current logic for the perf missing feature has a bug that it can
wrongly clear some modifiers like G or H.  Actually some PMUs don't
support any filtering or exclusion while others do.  But we check it as
a global feature.

For example, the cycles event can have 'G' modifier to enable it only in
the guest mode on x86.  When you don't run any VMs it'll return 0.

  # perf stat -a -e cycles:G sleep 1

    Performance counter stats for 'system wide':

                    0      cycles:G

          1.000721670 seconds time elapsed

But when it's used with other pmu events that don't support G modifier,
it'll be reset and return non-zero values.

  # perf stat -a -e cycles:G,msr/tsc/ sleep 1

    Performance counter stats for 'system wide':

          538,029,960      cycles:G
       16,924,010,738      msr/tsc/

          1.001815327 seconds time elapsed

This is because of the missing feature detection logic being global.
Add a hashmap to set pmu-specific exclude_host/guest features.

Committer notes:

Fix 'perf test python' by adding a stub for evsel__find_pmu() in
tools/perf/util/python.c, document that it is used so far only for the
above reasons so that if anybody needs this in the python binding
usecases, we can revisit this.

Reported-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@amd.com>
Link: http://lore.kernel.org/lkml/20211105205847.120950-1-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools/perf/util/evsel.c

@@ -1824,7 +1824,7 @@ static void evsel__disable_missing_features(struct evsel *evsel)
 		evsel->open_flags &= ~(unsigned long)PERF_FLAG_FD_CLOEXEC;
 	if (perf_missing_features.mmap2)
 		evsel->core.attr.mmap2 = 0;
-	if (perf_missing_features.exclude_guest)
+	if (evsel->pmu && evsel->pmu->missing_features.exclude_guest)
 		evsel->core.attr.exclude_guest = evsel->core.attr.exclude_host = 0;
 	if (perf_missing_features.lbr_flags)
 		evsel->core.attr.branch_sample_type &= ~(PERF_SAMPLE_BRANCH_NO_FLAGS |
@@ -1917,10 +1917,27 @@ bool evsel__detect_missing_features(struct evsel *evsel)
 		perf_missing_features.mmap2 = true;
 		pr_debug2_peo("switching off mmap2\n");
 		return true;
-	} else if (!perf_missing_features.exclude_guest &&
-		   (evsel->core.attr.exclude_guest || evsel->core.attr.exclude_host)) {
-		perf_missing_features.exclude_guest = true;
-		pr_debug2_peo("switching off exclude_guest, exclude_host\n");
+	} else if ((evsel->core.attr.exclude_guest || evsel->core.attr.exclude_host) &&
+		   (evsel->pmu == NULL || evsel->pmu->missing_features.exclude_guest)) {
+		if (evsel->pmu == NULL) {
+			evsel->pmu = evsel__find_pmu(evsel);
+			if (evsel->pmu)
+				evsel->pmu->missing_features.exclude_guest = true;
+			else {
+				/* we cannot find PMU, disable attrs now */
+				evsel->core.attr.exclude_host = false;
+				evsel->core.attr.exclude_guest = false;
+			}
+		}
+
+		if (evsel->exclude_GH) {
+			pr_debug2_peo("PMU has no exclude_host/guest support, bailing out\n");
+			return false;
+		}
+		if (!perf_missing_features.exclude_guest) {
+			perf_missing_features.exclude_guest = true;
+			pr_debug2_peo("switching off exclude_guest, exclude_host\n");
+		}
 		return true;
 	} else if (!perf_missing_features.sample_id_all) {
 		perf_missing_features.sample_id_all = true;

tools/perf/util/evsel.h

@@ -22,6 +22,7 @@ struct target;
 struct hashmap;
 struct bperf_leader_bpf;
 struct bperf_follower_bpf;
+struct perf_pmu;
 
 typedef int (evsel__sb_cb_t)(union perf_event *event, void *data);
 
@@ -153,6 +154,9 @@ struct evsel {
 	};
 	unsigned long		open_flags;
 	int			precise_ip_original;
+
+	/* for missing_features */
+	struct perf_pmu		*pmu;
 };
 
 struct perf_missing_features {

tools/perf/util/pmu.h

@@ -49,6 +49,10 @@ struct perf_pmu {
 	struct list_head caps;    /* HEAD struct perf_pmu_caps -> list */
 	struct list_head list;    /* ELEM */
 	struct list_head hybrid_list;
+
+	struct {
+		bool exclude_guest;
+	} missing_features;
 };
 
 extern struct perf_pmu perf_pmu__fake;

tools/perf/util/python.c

@@ -69,6 +69,18 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 {
 }
 
+/*
+ * This one is needed not to drag the PMU bandwagon, jevents generated
+ * pmu_sys_event_tables, etc and evsel__find_pmu() is used so far just for
+ * doing per PMU perf_event_attr.exclude_guest handling, not really needed, so
+ * far, for the perf python binding known usecases, revisit if this become
+ * necessary.
+ */
+struct perf_pmu *evsel__find_pmu(struct evsel *evsel __maybe_unused)
+{
+	return NULL;
+}
+
 /*
  * Add this one here not to drag util/metricgroup.c
  */