perf trace: Add --max-summary option

The --max-summary option is to limit the number of output lines for
syscall summary stats. The max applies to each entries like thread and
cgroups. For total summary, it will just print up to the given number.

For example,

$ sudo perf trace -as --max-summary 3 sleep 0.1

ThreadPoolServi (1011651), 114 events, 14.8%

syscall calls errors total min avg max stddev
(msec) (msec) (msec) (msec) (%)
--------------- -------- ------ -------- --------- --------- --------- ------
epoll_wait 38 0 95.589 0.000 2.515 11.153 28.98%
futex 9 0 0.040 0.002 0.004 0.014 28.63%
read 10 0 0.037 0.003 0.004 0.005 4.67%

sleep (1050529), 250 events, 32.4%

syscall calls errors total min avg max stddev
(msec) (msec) (msec) (msec) (%)
--------------- -------- ------ -------- --------- --------- --------- ------
clock_nanosleep 1 0 100.156 100.156 100.156 100.156 0.00%
execve 4 3 1.020 0.005 0.255 0.989 95.93%
openat 36 17 0.416 0.003 0.012 0.029 10.58%

...

And this is for per-cgroup summary using BPF.

$ sudo perf trace -as --max-summary 3 --summary-mode=cgroup --bpf-summary sleep 0.1

cgroup /user.slice/user-657345.slice/[e-mail]/session.slice/[e-mail], 12 events

syscall calls errors total min avg max stddev
(msec) (msec) (msec) (msec) (%)
--------------- -------- ------ -------- --------- --------- --------- ------
recvmsg 8 7 0.016 0.001 0.002 0.006 39.73%
ppoll 1 0 0.014 0.014 0.014 0.014 0.00%
write 2 0 0.010 0.002 0.005 0.008 61.02%

cgroup /user.slice/user-657345.slice/session-4.scope, 73 events

syscall calls errors total min avg max stddev
(msec) (msec) (msec) (msec) (%)
--------------- -------- ------ -------- --------- --------- --------- ------
epoll_wait 8 0 13.461 0.010 1.683 12.235 89.66%
ioctl 20 0 0.204 0.001 0.010 0.113 54.01%
writev 11 0 0.164 0.004 0.015 0.042 20.34%

Reviewed-by: Howard Chu <[e-mail]>
Signed-off-by: Namhyung Kim <[e-mail]>
Tested-by: Arnaldo Carvalho de Melo <[e-mail]>
Cc: Adrian Hunter <[e-mail]>
Cc: Ian Rogers <[e-mail]>
Cc: Ingo Molnar <[e-mail]>
Cc: Jiri Olsa <[e-mail]>
Cc: Kan Liang <[e-mail]>
Cc: Peter Zijlstra <[e-mail]>
Signed-off-by: Arnaldo Carvalho de Melo <[e-mail]>

tools/perf/Documentation/perf-trace.txt

···
 	works well with -s/--summary option where no argument information is
 	required.
 
+--max-summary=N::
+	Maximum number of lines in the summary mode.  Note that this applies to
+	each entry (thread or cgroup).
+
 
 PAGEFAULTS
 ----------

tools/perf/builtin-trace.c

···
 	unsigned int		max_stack;
 	unsigned int		min_stack;
 	enum trace_summary_mode	summary_mode;
+	int			max_summary;
 	int			raw_augmented_syscalls_args_size;
 	bool			raw_augmented_syscalls;
 	bool			fd_path_disabled;
···
 	if (!err) {
 		if (trace->summary) {
 			if (trace->summary_bpf)
-				trace_print_bpf_summary(trace->output);
+				trace_print_bpf_summary(trace->output, trace->max_summary);
 			else if (trace->summary_mode == SUMMARY__BY_TOTAL)
 				trace__fprintf_total_summary(trace, trace->output);
 			else
···
 				  struct hashmap *syscall_stats)
 {
 	size_t printed = 0;
+	int lines = 0;
 	struct syscall *sc;
 	struct syscall_entry *entries;
 
···
 						fprintf(fp, "\t\t\t\t%s: %d\n", perf_env__arch_strerrno(trace->host->env, e + 1), stats->errnos[e]);
 				}
 			}
+			lines++;
 		}
+
+		if (trace->max_summary && trace->max_summary <= lines)
+			break;
 	}
 
 	free(entries);
···
 	OPT_BOOLEAN(0, "force-btf", &trace.force_btf, "Prefer btf_dump general pretty printer"
 		       "to customized ones"),
 	OPT_BOOLEAN(0, "bpf-summary", &trace.summary_bpf, "Summary syscall stats in BPF"),
+	OPT_INTEGER(0, "max-summary", &trace.max_summary,
+		     "Max number of entries in the summary."),
 	OPTS_EVSWITCH(&trace.evswitch),
 	OPT_END()
 	};

tools/perf/util/bpf-trace-summary.c

···
 	return key1 == key2;
 }
 
-static int print_common_stats(struct syscall_data *data, FILE *fp)
+static int print_common_stats(struct syscall_data *data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 
-	for (int i = 0; i < data->nr_nodes; i++) {
+	if (max_summary == 0 || max_summary > data->nr_nodes)
+		max_summary = data->nr_nodes;
+
+	for (int i = 0; i < max_summary; i++) {
 		struct syscall_node *node = &data->nodes[i];
 		struct syscall_stats *stat = &node->stats;
 		double total = (double)(stat->total_time) / NSEC_PER_MSEC;
···
 	return 0;
 }
 
-static int print_thread_stat(struct syscall_data *data, FILE *fp)
+static int print_thread_stat(struct syscall_data *data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 
···
 	printed += fprintf(fp, "                                     (msec)    (msec)    (msec)    (msec)        (%%)\n");
 	printed += fprintf(fp, "   --------------- --------  ------ -------- --------- --------- ---------     ------\n");
 
-	printed += print_common_stats(data, fp);
+	printed += print_common_stats(data, max_summary, fp);
 	printed += fprintf(fp, "\n\n");
 
 	return printed;
 }
 
-static int print_thread_stats(struct syscall_data **data, int nr_data, FILE *fp)
+static int print_thread_stats(struct syscall_data **data, int nr_data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 
 	for (int i = 0; i < nr_data; i++)
-		printed += print_thread_stat(data[i], fp);
+		printed += print_thread_stat(data[i], max_summary, fp);
 
 	return printed;
 }
···
 	return 0;
 }
 
-static int print_total_stats(struct syscall_data **data, int nr_data, FILE *fp)
+static int print_total_stats(struct syscall_data **data, int nr_data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 	int nr_events = 0;
···
 	printed += fprintf(fp, "                                     (msec)    (msec)    (msec)    (msec)        (%%)\n");
 	printed += fprintf(fp, "   --------------- --------  ------ -------- --------- --------- ---------     ------\n");
 
-	for (int i = 0; i < nr_data; i++)
-		printed += print_common_stats(data[i], fp);
+	if (max_summary == 0 || max_summary > nr_data)
+		max_summary = nr_data;
+
+	for (int i = 0; i < max_summary; i++)
+		printed += print_common_stats(data[i], max_summary, fp);
 
 	printed += fprintf(fp, "\n\n");
 	return printed;
···
 	return 0;
 }
 
-static int print_cgroup_stat(struct syscall_data *data, FILE *fp)
+static int print_cgroup_stat(struct syscall_data *data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 	struct cgroup *cgrp = __cgroup__find(&cgroups, data->key);
···
 	printed += fprintf(fp, "                                     (msec)    (msec)    (msec)    (msec)        (%%)\n");
 	printed += fprintf(fp, "   --------------- --------  ------ -------- --------- --------- ---------     ------\n");
 
-	printed += print_common_stats(data, fp);
+	printed += print_common_stats(data, max_summary, fp);
 	printed += fprintf(fp, "\n\n");
 
 	return printed;
 }
 
-static int print_cgroup_stats(struct syscall_data **data, int nr_data, FILE *fp)
+static int print_cgroup_stats(struct syscall_data **data, int nr_data, int max_summary, FILE *fp)
 {
 	int printed = 0;
 
 	for (int i = 0; i < nr_data; i++)
-		printed += print_cgroup_stat(data[i], fp);
+		printed += print_cgroup_stat(data[i], max_summary, fp);
 
 	return printed;
 }
 
-int trace_print_bpf_summary(FILE *fp)
+int trace_print_bpf_summary(FILE *fp, int max_summary)
 {
 	struct bpf_map *map = skel->maps.syscall_stats_map;
 	struct syscall_key *prev_key, key;
···
 
 	switch (skel->rodata->aggr_mode) {
 	case SYSCALL_AGGR_THREAD:
-		printed += print_thread_stats(data, nr_data, fp);
+		printed += print_thread_stats(data, nr_data, max_summary, fp);
 		break;
 	case SYSCALL_AGGR_CPU:
-		printed += print_total_stats(data, nr_data, fp);
+		printed += print_total_stats(data, nr_data, max_summary, fp);
 		break;
 	case SYSCALL_AGGR_CGROUP:
-		printed += print_cgroup_stats(data, nr_data, fp);
+		printed += print_cgroup_stats(data, nr_data, max_summary, fp);
 		break;
 	default:
 		break;

tools/perf/util/trace.h

···
 int trace_prepare_bpf_summary(enum trace_summary_mode mode);
 void trace_start_bpf_summary(void);
 void trace_end_bpf_summary(void);
-int trace_print_bpf_summary(FILE *fp);
+int trace_print_bpf_summary(FILE *fp, int max_summary);
 void trace_cleanup_bpf_summary(void);
 
 #else /* !HAVE_BPF_SKEL */
···
 }
 static inline void trace_start_bpf_summary(void) {}
 static inline void trace_end_bpf_summary(void) {}
-static inline int trace_print_bpf_summary(FILE *fp __maybe_unused)
+static inline int trace_print_bpf_summary(FILE *fp __maybe_unused, int max_summary __maybe_unused)
 {
 	return 0;
 }