perf stat: Fix counting when initial delay configured

When creating counters with initial delay configured, the enable_on_exec
field is not set. So we need to enable the counters later. The problem
is, when a workload is specified the target__none() is true. So we also
need to check stat_config.initial_delay.

In this change, we add a new field 'initial_delay' for struct target
which could be shared by other subcommands. And define
target__enable_on_exec() which returns whether enable_on_exec should be
set on normal cases.

Before this fix the event is not counted:

  $ ./perf stat -e instructions -D 100 sleep 2
  Events disabled
  Events enabled

   Performance counter stats for 'sleep 2':

       <not counted>      instructions

         1.901661124 seconds time elapsed

         0.001602000 seconds user
         0.000000000 seconds sys

After fix it works:

  $ ./perf stat -e instructions -D 100 sleep 2
  Events disabled
  Events enabled

   Performance counter stats for 'sleep 2':

             404,214      instructions

         1.901743475 seconds time elapsed

         0.001617000 seconds user
         0.000000000 seconds sys

Fixes: c587e77e ("perf stat: Do not delay the workload with --delay")
Signed-off-by: Changbin Du <changbin.du@huawei.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Hui Wang <hw.huiwang@huawei.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20230302031146.2801588-2-changbin.du@huawei.comSigned-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools/perf/builtin-stat.c

@@ -539,12 +539,7 @@ static int enable_counters(void)
 			return err;
 	}
 
-	/*
-	 * We need to enable counters only if:
-	 * - we don't have tracee (attaching to task or cpu)
-	 * - we have initial delay configured
-	 */
-	if (!target__none(&target)) {
+	if (!target__enable_on_exec(&target)) {
 		if (!all_counters_use_bpf)
 			evlist__enable(evsel_list);
 	}
@@ -914,7 +909,7 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx)
 			return err;
 	}
 
-	if (stat_config.initial_delay) {
+	if (target.initial_delay) {
 		pr_info(EVLIST_DISABLED_MSG);
 	} else {
 		err = enable_counters();
@@ -926,8 +921,8 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx)
 	if (forks)
 		evlist__start_workload(evsel_list);
 
-	if (stat_config.initial_delay > 0) {
-		usleep(stat_config.initial_delay * USEC_PER_MSEC);
+	if (target.initial_delay > 0) {
+		usleep(target.initial_delay * USEC_PER_MSEC);
 		err = enable_counters();
 		if (err)
 			return -1;
@@ -1248,7 +1243,7 @@ static struct option stat_options[] = {
 		     "aggregate counts per thread", AGGR_THREAD),
 	OPT_SET_UINT(0, "per-node", &stat_config.aggr_mode,
 		     "aggregate counts per numa node", AGGR_NODE),
-	OPT_INTEGER('D', "delay", &stat_config.initial_delay,
+	OPT_INTEGER('D', "delay", &target.initial_delay,
 		    "ms to wait before starting measurement after program start (-1: start with events disabled)"),
 	OPT_CALLBACK_NOOPT(0, "metric-only", &stat_config.metric_only, NULL,
 			"Only print computed metrics. No raw values", enable_metric_only),

tools/perf/util/stat.c

@@ -842,11 +842,7 @@ int create_perf_stat_counter(struct evsel *evsel,
 	if (evsel__is_group_leader(evsel)) {
 		attr->disabled = 1;
 
-		/*
-		 * In case of initial_delay we enable tracee
-		 * events manually.
-		 */
-		if (target__none(target) && !config->initial_delay)
+		if (target__enable_on_exec(target))
 			attr->enable_on_exec = 1;
 	}
 

tools/perf/util/stat.h

@@ -166,7 +166,6 @@ struct perf_stat_config {
 	FILE			*output;
 	unsigned int		 interval;
 	unsigned int		 timeout;
-	int			 initial_delay;
 	unsigned int		 unit_width;
 	unsigned int		 metric_only_len;
 	int			 times;

tools/perf/util/target.h

@@ -18,6 +18,7 @@ struct target {
 	bool	     per_thread;
 	bool	     use_bpf;
 	bool	     hybrid;
+	int	     initial_delay;
 	const char   *attr_map;
 };
 
@@ -72,6 +73,17 @@ static inline bool target__none(struct target *target)
 	return !target__has_task(target) && !target__has_cpu(target);
 }
 
+static inline bool target__enable_on_exec(struct target *target)
+{
+	/*
+	 * Normally enable_on_exec should be set if:
+	 *  1) The tracee process is forked (not attaching to existed task or cpu).
+	 *  2) And initial_delay is not configured.
+	 * Otherwise, we enable tracee events manually.
+	 */
+	return target__none(target) && !target->initial_delay;
+}
+
 static inline bool target__has_per_thread(struct target *target)
 {
 	return target->system_wide && target->per_thread;